From da at ski.org  Fri Nov  5 15:59:35 1999
From: da at ski.org (David Ascher)
Date: Fri, 5 Nov 1999 12:59:35 -0800 (Pacific Standard Time)
Subject: IPC8: SouthWest Airlines Internet specials
In-Reply-To: <14371.16525.555599.961527@dolphin.mojam.com>
Message-ID: <Pine.WNT.4.04.9911051248370.260-100000@rigoletto.ski.org>

On Fri, 5 Nov 1999, Skip Montanaro wrote:

> 
>     Jeremy> Everyone planning to fly to the conference should consider
>     Jeremy> flying through BWI.  I have flown from BWI to San Jose and San
>     Jeremy> Francisco a few times this year, each time for about $200.
>     Jeremy> Flights through BWI are often cheaper than flights from Dulles
>     Jeremy> or National.
> 
> If you're on a budget and might like to get away without renting a car,
> don't forget to figure in the cost of ground transport into the total
> equation.  For IPC1 I flew into BWI and found the metro didn't go there.
> Not real convenient.  At least with National you've got a Metro station
> nearby.  

FYI: Supershuttle from BWI to Alexandria VA is $35 according to the
supershuttle operator.  Cheaper for more than one person.  We might be
able to get group discounts as well...

There are also limos for about $60/car (4 people).

I've updated the GroupMind site, so that if folks want to coordinate limo
rides, room sharing, etc., it shouldn't be too hard:

http://starship.python.net/crew/scharf/GroupMind/gromi.cgi?req=show&file=ConferenceStuff.gromi

  (see "TravelPlans" page)

--david

Also FYI: http://www.bwiairport.com/original/ground_nv.html






From mwh21 at cam.ac.uk  Thu Nov  4 07:17:14 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 04 Nov 1999 12:17:14 +0000
Subject: return boolean from functions
References: <7vrq9b$hmr$1@nnrp1.deja.com>
Message-ID: <m3iu3ia1th.fsf@atrus.jesus.cam.ac.uk>

a_olme at my-deja.com writes:

> Hello.
> 
> Is it possible to return booelan from functions other than return 1==1
> for true or return 1==2 for false?

Well, you could return 1 for true and 0 for false:

>>> print 1==1, type(1==1), 1==0, type(1==0)
1 <type 'int'> 0 <type 'int'>

Python does not have a distinct boolean type, if that's what you're
asking. Empty lists, empty strings, 0, 0L and 0.0 are false,
everything else is true (except for instances of classes that have
__nozero__ methods, which get to decide for themselves).

Chapter and verse:

http://www.python.org/doc/current/ref/lambda.html#Booleans

HTH,
Michael




From gerrit.holl at pobox.com  Mon Nov  1 14:16:36 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Mon, 1 Nov 1999 20:16:36 +0100
Subject: Ultimate newbie question? How to execute DOS commands via Python.
In-Reply-To: <7309F1998D14D3119DD80008C7C9CA80631132@GIBRALTAR>
References: <7309F1998D14D3119DD80008C7C9CA80631132@GIBRALTAR>
Message-ID: <19991101201635.A29736@optiplex.palga.uucp>

Gabe Newcomb wrote:
> Hi all:
> 	My job here at work would be so much easier if  I could use Python
> to run a series of DOS commands (I'd rather do this than write a batch file
> for a number of reasons). Any pointers/examples on this would extremely
> helpful!

It's in the documentation:

os module:
  system (command)
      Execute the command (a string) in a subshell. This is implemented by
      calling the Standard C function system(), and has the same limitations.
      Changes to posix.environ, sys.stdin etc. are not reflected in the
      environment of the executed command. The return value is the exit status
      of the process encoded in the format specified for wait().

I don't know how to get the output on DOS systems: os.pipe() and sisters are
not available on your platform, but this may be well enough.

regards,
Gerrit.
-- 
linuxgames.nl.linux.org	 All about games on Linux in Dutch (under construction).
www.nl.linux.org	       The Dutch resource for Dutch Linux documentation.
www.nl.linux.org/discoverb		    Learn foreign words and definitions.
www.nl.linux.org/~gerrit/asperger	 I'm an Asperger, go here for more info.
www.nl.linux.org/~gerrit			    About me and pictures of me.



From gek8365 at garnet.acns.fsu.edu  Tue Nov  2 10:47:10 1999
From: gek8365 at garnet.acns.fsu.edu (Glenn Kidd)
Date: Tue, 2 Nov 1999 10:47:10 -0500
Subject: Template for CGI
Message-ID: <7vn0sk$7cp$1@news.fsu.edu>

I am a newbie to Python and I have a question.  I want a CGI script that
will read a file template and replace certain words with variables.  Ii
tried putting the variable names in the file but it is read as palin text.
Is there a way to embed variables into a teplate or a replace function in
Python?  Thank you for any help you give.


Glenn Kidd





From ryazonjr at home.com  Mon Nov  8 23:44:02 1999
From: ryazonjr at home.com (bob yazon)
Date: Tue, 09 Nov 1999 04:44:02 GMT
Subject: python: something completely different
References: <3820C0B8.62B2@iowegian.com> <38249ED2.BB7EA1DE@znet.com> <8028jv$7cf$1@hub.org> <38259F18.437@iowegian.com> <38278B5F.1AFF953C@python.net>
Message-ID: <mENV3.6280$ra7.129744@news1.wwck1.ri.home.com>

The orig post got me interested in the stuff. I just finished installing it
+ win32ext, numeric, and gnuplot.  I can get pretty good graphs using
gnuplot, I think I'll start using python instead of Mlab.

--ry

Thomas A. Bryan <tbryan at python.net> wrote in message
news:38278B5F.1AFF953C at python.net...
> we're translating to a Python with Numeric Python compiled).  It
> might be a fun project, but it probably wouldn't move too many
> people away from Matlab until they could get colored, labeled,
> scaled plots from Python.  Most people would also have to be willing






From a_olme at my-deja.com  Thu Nov  4 06:22:20 1999
From: a_olme at my-deja.com (a_olme at my-deja.com)
Date: Thu, 04 Nov 1999 11:22:20 GMT
Subject: return boolean from functions
Message-ID: <7vrq9b$hmr$1@nnrp1.deja.com>

Hello.

Is it possible to return booelan from functions other than return 1==1
for true or return 1==2 for false?

//Anders Olme


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



From mal at lemburg.com  Mon Nov 15 04:37:16 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Mon, 15 Nov 1999 10:37:16 +0100
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us>
Message-ID: <382FD44C.C85BCE7F@lemburg.com>

"Andrew M. Kuchling" wrote:
> 
> http://www.vex.net/~x/parnassus/ is *very* nice, and it might be the
> only solution required.  Apparently a good deal of work has been put
> into seeding the initial database, and that's important, because it
> makes the index immediately useful.  Yet I hadn't heard about it until
> now; is the site still beta?  Did I miss a c.l.py.announce article
> about it?

Could you put a link to it from www.python.org/search ?

It does indeed look *very* nice; cheers to the author(s) :-)

-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                    46 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/





From pinard at iro.umontreal.ca  Tue Nov  2 09:09:54 1999
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 02 Nov 1999 09:09:54 -0500
Subject: Caps in doc strings (was: Re: Default indentation of 4 columns)
In-Reply-To: Paul Foley's message of "02 Nov 1999 22:34:24 +1300"
References: <oq904sc08y.fsf@titan.progiciels-bpi.ca> <86emeko88f.fsf@g.local> <oqeme9hg66.fsf@titan.progiciels-bpi.ca> <m23dupnsnz.fsf@mycroft.actrix.gen.nz>
Message-ID: <oqwvs1nfwt.fsf_-_@titan.progiciels-bpi.ca>

Paul Foley <mycroft at actrix.gen.nz> ?crit:

[about using capitals or not in documentation strings]

> Interestingly, I use both conventions, but for different reasons (I
> didn't even know that the use of capitals _was_ a GNU convention; did know
> Knuth uses |vertical bars| for literals, but hadn't even considered that).

  [I wanted to include the Python list in this reply, and for doing so, took
  the liberty of quoting most of your message, which I found interesting
  and instructive, to provide enough context.]

In Texinfo, @var{xyz} is for documenting variables, and is rendered
with slanted fonts on a laser printer.  When converting to Info (that is,
ASCII), there is no slanted fonts, and capitals are used instead.  In fact,
exactly how various @style{...} are represented in ASCII is well documented.

The informal convention for doc strings is letting ourselves think Texinfo
(@code{...} for code elements, @samp{...} for string fragments, @var{...}
for variables or non-terminal symbols, @emph{...} for emphasis, etc.) and
just doing the Info rendering "by hand", following the standards set by
the Texinfo to Info conversion.

> I use capitals because that's the common case in Common Lisp; I use
> vertical bars for lowercase symbols for the same reason (Common Lisp uses
> vertical bars to read in a case-preserving manner).  Hence, I use vertical
> bars for Python variables and functions, because they're case-dependent and
> case-preserving, and capitals in languages that are either case-insensitive
> (e.g., Pascal) or that standardize on upper-case (e.g., Lisp).

I thought that, in Python, usual local variables, and formal arguments among
them, are to be conventionally set in lower case anyway, so the mixed case
difficulty does not apply in actual practice, even the language itself would
allow more variance.  No doubt that a coding style is a set of restrictions
over what a language permits, that we choose to follow for various reasons.

There is a lot of commonality (or at least, very little clash) between
Guido's recommendations and GNU standards about the naming convention for
identifiers, and it was part of my pleasure, reading the details of Python
coding style, to see that I had so little to change in my long lasting
habits.  (The main change was to learn how to use a bit less whitespace
at some specific places, like between a function and its argument list,
or around the `=' introducing keyword arguments -- and surprisingly,
it was very easy to get some speed on this).  I consider that Guido's
recommendations and GNU standards enlighten each other to a good extent.

Besides Emacs LISP, I have little recent exposure to various dialects
of LISP.  What I've seen from others' Scheme code (but that is not
much) seems to favour lower-case as well, and capitals for systemic
constants, nothing unusual for me.  Scheme ignores case in identifiers,
but this never precluded me to write exactly as if case was significant.
DSSSL, if I remember well, is a case-sensitive implementation of Scheme.
(So is Guile, the last I heard, but since Guile was not meant to be very
respectful of standards when initially designed, I return the favour by
not being very respectful of Guile; what Guile does do not matter much. :-)

> I think the "use caps" standard is a pretty bad idea in a case-sensitive
> case-preserving language like Python, since it doesn't let you distinguish
> between names that the same in all but case (common to make the distinction
> between class names and variable names, in Python).

Python and C (and Emacs LISP, for that matter) are quite similar in what they
permit for capitalisation, and what people actually use while writing these.
I quite understand what you mean by "use caps" standard being a pretty
bad idea, and I remember feeling uncomfortable when I started to use it,
thinking about those constants we always write in all capitals, before even
speaking of them in doc strings, and all the confusion that might result.

But I followed the capitalisation convention in doc strings for more than
a dozen years, being often quite active at programming, and I can ensure
you from experience that my fears were not justified.  That convention is
quite workable, it does not raise any real problem on the wide average.
I'm rather convinced, by now, that it is a legible way to write doc strings.
If we simultaneously follow the rest of naming conventions, that is! :-)

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





From gmcm at hypernet.com  Tue Nov  9 15:32:15 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: 9 Nov 1999 14:32:15 -0600
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Nov  9)
Message-ID: <19C38A3431A7057C.8DB369AEF127817A.5E7393A58AA59D18@lp.airnews.net>

This week on comp.lang.python we find:

    Evan Simpson's one page description of the anatomy of Zope:
        http://www.deja.com/getdoc.xp?AN=544506548

    Paul Duffin's speculation about how "stubbed" Tcl might be
    used by Tkinter:
        http://www.deja.com/getdoc.xp?AN=545858112

    The __str__ vs __repr__ thread rages on.
    Guido explains the rationale:     
        http://www.deja.com/getdoc.xp?AN=544209311
    Tim clarifies the problem:
        http://www.deja.com/getdoc.xp?AN=544718491
    Bernhard Herzog makes an interesting proposal:      
        http://www.deja.com/getdoc.xp?AN=544360351
    while others split on whether "eval(repr(obj)) == obj" is a
    valuable property or a wart.

    Jean-Claude Wippler announces a stable beta of his Python
    wrappers for MetaKit (a cross-platform embedded database
    with intriguing properties):
        http://www.equi4.com/metakit/python.html

    Registration for IPC8 is now online:
        http://www.python.org/workshops/2000-01/registration.html
    and prompts some discussion of travel arrangements.

    Two new Python books get rave reviews:
      David Beazley's Python Essential Reference
    and
      Fredrik Lundh's ebook:
      (the eff-bot guide to) The Standard Python Library
    (although technical / platform problems create some
    controversy for the eff-bot!).
        http://www.python.org/psa/bookstore/

    Greg Ewing compares Python and Ruby and makes some
    interesting observations: 
      on extension writing interface:
        http://www.deja.com/getdoc.xp?AN=543792249
      on the object model:
        http://www.deja.com/getdoc.xp?AN=544255921

    Python scores points on rec.humor.funny:
        http://www.netfunny.com/rhf/jokes/99/Nov/perl.html


=========================================================================
Everything you want is probably one or two clicks away in these pages:

      Python.org's Python Language Website is the center of Pythonia
          http://www.python.org

      Much of Python's real work takes place on Special-Interest Group
          mailing lists http://www.python.org/sigs/

      Python Consortium emerges as an independent nexus of activity
          http://www.python.org/consortium

      Python To-Do List anticipates some of Python's future direction
          http://www.python.org/cgi-bin/todo.py

      Python Journal is at work on its second issue
          http://www.pythonjournal.com

      Links2Go is a new semi-automated link collection; it's impressive
      what AI can generate
          http://www.links2go.com/search?search=python

    Archive probing trick of the trade:
        http://www.dejanews.com/dnquery.xp?QRY=&DBS=2&ST=PS&defaultOp=AND&LNG=ALL&format=threaded&showsort=date&maxhits=100&groups=comp.lang.python


Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://purl.org/thecliff/python/url.html
or
  http://www.dejanews.com/dnquery.xp?QRY=~g%20comp.lang.python%20Python-URL%21


Suggestions/corrections for next week's posting are always welcome.
  http://www.egroups.com/list/python-url-leads/

To receive a new issue of this posting in e-mail each Monday morning,
ask <claird at neosoft.com> to subscribe.  Mention "Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.
-- 

Cameron Laird           http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX



From phd at phd.russ.ru  Wed Nov 24 06:59:29 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Wed, 24 Nov 1999 11:59:29 +0000 (GMT)
Subject: Python 1.5.2: time.strptime() may be broken
In-Reply-To: <38344cd9$0$232@nntp1.ba.best.com>
Message-ID: <Pine.LNX.4.21.9911241158390.1298-100000@fep132.fep.ru>

On 18 Nov 1999, David S. Harrison wrote:
> When I run this in Python 1.5.2 on Solaris 2.5.1 I get

   Few weeks ago I posted almost the same problem on the same Solaris
platform. Still cannot resolve the issue...

Oleg.
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From mlinnen at usa.net  Mon Nov  1 22:46:01 1999
From: mlinnen at usa.net (Mike Linnen)
Date: Tue, 02 Nov 1999 03:46:01 GMT
Subject: parallel port programming using Python?
References: <19991014.22042400@mis.configured.host> <KimO3.9837$pL5.20871@news.rdc1.az.home.com> <7vc1tg$5477@eccws12.dearborn.ford.com>
Message-ID: <Z7tT3.3398$gX.22823@news.rdc1.az.home.com>

Most of my robotic projects have revolved around 8 bit microcontrolers and
various languages like Basic, Forth and assembly.  I was always running into
problems like lack of floating point arithmetic, or running out of memory.
So I recently moved to Windows 98 platform on a laptop with Visual Basic as
the language. The laptop would communicate through the RS232 port to a slave
computer.  The slave computer handled all the time specific stuff like PWM
to the motors,  collision avoidance ect.. Visual Basic was used for the high
level stuff.

I wanted to try some Object Oriented techniques and I ran across Python.
Since it is an interpreted language I thought it would be easy to prototype
new ideas. So far it has worked very well for me. I do not have a full robot
functioning with Linux and Python yet, but I did write some C code to access
a Vector 2X compass on the parallel port.  I then extended the Python
language with my new V2X module.  I used the demo/extend/xxmodule.C as an
example for my module.  Now I can simply import my new module inside Python
and read the direction my robot is facing with a couple of commands.  All
the high level mapping/ navigational planning can be done with Python.  This
seems like a cool way to go as far as I am concerned.  Modeling Behaviors in
Python using OOP would be really neat.  Extending the robots functionality
should become easier also.

I am not a real C programmer myself but I was able to port my Visual Basic
V2X code over to C fairly easy.  I got it working on a Linux box without
Python first.  Then I merged Python and my working C code together.

Currently I am researching what Single Board Computer will best suit my
needs.  I will be using Linux, Python, and C together to achieve almost all
of my I/O.  I suspect I might have to go to RTLINUX to achieve this.

I am also interested in your progress with Linux and your Bot.  I have only
used DC motors and PWM myself but have thought of going to steppers for more
accuracy.


Mark Ricketts <mricket2 at ford.com> wrote in message
news:7vc1tg$5477 at eccws12.dearborn.ford.com...
> I'm kind of a novice when it comes to c programming, I've used basic on
the
> pc but since I'm acquaninting myself with linux I figured the best way to
> learn a new language is to attempt an actual useful project.  I'd like to
> build a little xyz robot using stepper motors for motion control that will
> be operated via the parallel port.  I've already got the thing working in
> the DOS/Windows world, but I am so into the whole linux philosophy that I
> want to redo it using Python in Linux.  Python is said to be a very easy
> language to pick-up but so far I haven't found any language routines that
> will allow me to access the parallel port for this type of control.  So
Mike
> I'm very interested in what you are doing, maybe you could expand on what
> your project is all about, and how it's progressing.
>
>
> ------
> msdr at sprint.ca
>
>
> Mike Linnen <mlinnen at usa.net> wrote in message
> news:KimO3.9837$pL5.20871 at news.rdc1.az.home.com...
> > I am in the process of rolling my own extension to python that allows
> > parrallel port access.  If you know C it is not to hard to extend
pyhtons
> > functionallity.
> >
> >
> > Mark Rickets <markr at tfz.net> wrote in message
> > news:19991014.22042400 at mis.configured.host...
> > I want to use python for data acquisition from the parallel port.  I
> > am using Mandrake-Linux as the operating system  on a PC if that makes
> > a difference.  Are there any modules available in Python that allow
> > one to access the system parallel port for this type of function?
> >
> > msdr at sprint.ca
> >
> >
> >
> >
> >
>
>





From gmcm at hypernet.com  Tue Nov 30 16:22:26 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: 30 Nov 1999 15:22:26 -0600
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Nov 30)
Message-ID: <73E3D5EED5B97534.F05536BF851AFF0C.39137DDED1DD80F5@lp.airnews.net>

Noteworthy items this week from Pythondom:

    Michael Stroeder announces pyCA-0.6.2 - a set of
    scripts and CGI-BIN programs written in Python for
    setting up and running a certificate authority using
    OpenSSL.
        http://www.deja.com/=dnc/getdoc.xp?AN=553266414
 
    Sean Reifschneider announces sockserv, a SMTP/NNTP/POP-
    like server module that makes it easy to create command /
    response type servers.
        http://www.deja.com/=dnc/getdoc.xp?AN=553266413

    Tony Ibbs announces v0.3 for his metalanguage for
    mxTextTools
        http://www.deja.com/=dnc/getdoc.xp?AN=553266415
    [Note: mxTextTools features an unusual and extremely
    fast parsing engine with an incomprehensible API; Tony
    has been slowly chipping away at making the interface
    non-hostile for a year or so.]

    Mike Fletcher posts mandelbrot.py 
        http://www.deja.com/=dnc/getdoc.xp?AN=554408930
    which shows off both wxPython and the Numeric extensions.

    The subject of automatic garbage collection returns, this
    time with a practical cast: Fredrik Lundh on how it would
    be helpful:
        http://www.deja.com/=dnc/getdoc.xp?AN=553393467
    James Logajan on why he prefers the current approach: 
        http://www.deja.com/=dnc/getdoc.xp?AN=553396826
    And finally, Bijan Parsia gives pointers to some modern
    GC implementations that should ease James' concerns.
        http://www.deja.com/=dnc/getdoc.xp?AN=553786287

    Tim Peters exposes his impractical bent, and shows that an
    apparently simple "lottery" game isn't so simple:
        http://www.deja.com/=dnc/getdoc.xp?AN=552877307
    and
        http://www.deja.com/=dnc/getdoc.xp?AN=553253462


========================================================================

Everything you want is probably one or two clicks away in these pages:

      Python.org's Python Language Website is the center of Pythonia
          http://www.python.org

      Much of Python's real work takes place on Special-Interest Group
          mailing lists http://www.python.org/sigs/

      Python Consortium emerges as an independent nexus of activity
          http://www.python.org/consortium

      Python To-Do List anticipates some of Python's future direction
          http://www.python.org/cgi-bin/todo.py

      Python Journal is at work on its second issue
          http://www.pythonjournal.com

      Links2Go is a new semi-automated link collection; it's impressive
      what AI can generate
          http://www.links2go.com/search?search=python

      Archive probing trick of the trade:
          http://www.dejanews.com/dnquery.xp?QRY=&DBS=2&ST=PS&defaultOp=AND&LNG=ALL&format=threaded&showsort=date&maxhits=100&groups=comp.lang.python


Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://purl.org/thecliff/python/url.html
or
  http://www.dejanews.com/dnquery.xp?QRY=~g%20comp.lang.python%20Python-URL%21


Suggestions/corrections for next week's posting are always welcome.
  http://www.egroups.com/list/python-url-leads/

To receive a new issue of this posting in e-mail each Monday morning,
ask <claird at neosoft.com> to subscribe.  Mention "Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.
-- 

Cameron Laird           http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX



From botnivick at hotmail.com  Thu Nov 25 16:29:39 1999
From: botnivick at hotmail.com (Jim Bishop)
Date: Thu, 25 Nov 1999 21:29:39 GMT
Subject: Newbie Question:  Generating Timer Events wit Tkinter
Message-ID: <7Th%3.30$Zh.1100@wdc-read-01.qwest.net>

begin 644 Happy99.exe
M35I0``(````$``\`__\``+@`````````0``:````````````````````````
M``````````````````````$``+H0``X?M`G-(;@!3,TAD)!4:&ES('!R;V=R
M86T@;75S="!B92!R=6X@=6YD97(@5VEN,S(-"B0W````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M`````````````````````````````````````````%!%``!,`00`GR77C@``
M````````X`". at 0L!`AD`"@```!8```````````$````!`````@```$`````!
M```"```!``````````,`"@`````````%```$`````````@``````$```(```
M```0```0````````$``````````````````#`$`#````````````````````
M``````````````````0`:`$`````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````0T]$10``````$``````!```*````!@``````````````````
M(```8$1!5$$``````!```````@``$````!```````````````````$```,`N
M:61A=&$````0``````,```0````@``````````````````!```#`+G)E;&]C
M````$``````$```"````)```````````````````0```4```````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````:`!X``!J0.C6"```A<`/A!T&``!0O^L.0@"K!6!M``"K!<@```"K
M6%!04%^XE````*OHMP@``%Z#QA"M at _@`#X3L!0``OLD-0@!67[F5````K/;0
MJN+Z:,@```#_->\.0@#HC0@``(7`#X2W!0``H_<.0@!HR````/\U\PY"`&H`
MZ%8(``"%P`^$F`4``(LU\PY"``/P@^X%K"3?/$%U"L<%B@]"`/____^^(PY"
M`(L][PY"``,]]PY"`+D)````\Z1J`?\U[PY"`/\U\PY"`.CO!P``O@``0@"+
M/>L.0@"M/45.1"!T%3U:15)/=`7WT*OK[*V+R#/`\ZOKXXO/*PWK#D(`B0W[
M#D(`OAH.0@"+/>\.0@`#/?<.0@"Y"0```/.D,\!0:(````!J`E!0:````$#_
M->\.0@#HNP<``$`/A.L$``!(H_\.0@!J`&C[#D(`_S7[#D(`_S7K#D(`_S7_
M#D(`Z$('``"%P`^$M`0``+X-#D(`BSWO#D(``SWW#D(`N0T```#SI(LU[PY"
M`(L]\PY"`(L-]PY"`(/!"?.DN'-K80"K:@'_-?,.0@#_->\.0@#H"@<``#/`
M4&B`````:@-04&@```#`_S7O#D(`Z"0'``!`=5(SP/\UZPY"`&@'#T(`4&@_
M`!\`4%!0:"P.0@!H`@``@.@@!P``N`@```!0N",.0@!`4&H!:@!0_S4'#T(`
MZ/T&``#_-0</0@#H_ at 8``.G[`P``2*->#D(`,\!04%!J!%#_-5X.0@#HI`8`
M`(7`#X3/`P``HV8.0@`SP%!04&H&_S5F#D(`Z*D&``"%P`^$I0,``*-J#D(`
MB_!F at 3Y-6@^%DP,``(!^$GH/A(D#``#&1A)Z`W8\9H$^4$4/A7<#``")-7(.
M0@!FBT8&9J-V#D(`,\EFBPUV#D(`9HM&%&:C>`Y"`(O>@\,8,\!F`P5X#D(`
M`]B+`STN=&5X=",]+F5D870//2YD871T68/#*$EUX^M>BT,,*T,4HWH.0@#K
MY/=#)"```&`/A"P#``"!2R0```"`B1V>#D(`BT,0BWL(*\<]R@````^"#`,`
M`(M##(M3%"O"HWX.0@`#UXD5D at Y"`.N9BT,,*T,4HX(.0@#KFK^&#D(`BQ5Z
M#D(`BUYXBS5J#D(`*]H#WHM#'"O"`\:KBT,@*\(#QJN+0R0KP@/&JXM+&#/2
MBS6*#D(`QP6B#D(``````(L>*QUZ#D(``QUJ#D(`BP,]8V]N;G0@/7-E;F1T
M8D*#Q at 1)==N#/:(.0@`"#X5Q`@``Z9(```"#PP2+`SUE8W0`==M25HL=C at Y"
M`-'B`]HSP&:+`XLUA at Y"`,'@`@/PBP:CE at Y"`*&2#D(``P5^#D(`@\``B0;_
M!:(.0@!>6NN>@\,$B@,\`'654E:+'8X.0@#1X@/:,\!FBP.+-88.0@#!X`(#
M\(L&HYH.0@"AD at Y"``,%?@Y"`(/`1XD&_P6B#D(`7EKI5?___XLUG at Y"`(%&
M",H```!HJ at Y"`.A0!```A<`/A)H!``"CI at Y"`&BW#D(`_S6F#D(`Z#\$``"%
MP`^$?0$``*/?#D(`:,0.0@#_-:8.0@#H(@0``(7`#X1@`0``H^,.0@!HT`Y"
M`/\UI at Y"`.@%!```A<`/A$,!``"CYPY"`(L]D at Y"``,]:@Y"`.C*````G&#H
M`````%^!Q[T```"+7"0LBD,#/!EU"(M$)"BJ1^L*/'=U&T>+1"0HJN@(````
M4VMA+F1L;`"X_______0JV&=Z0````"<8.@`````7H/&=F:MBUPD*#KC=!`Z
MPW0"ZUOH#P```&UA:6P`Z`4```!N97=S`*U0N/______T(7`=#K_T#P!=#1F
MD^@`````7H/&-%9?,\"`^TYU"D>JK#P`=1E&ZPV`^TUU$:I'1JP\`'4)K5"X
M_______089WI`````````````%ZYR@```/.DH=\.0@")AV____^AYPY"`(E'
MKZ'C#D(`B4?MBQ5^#D(`H9(.0@`#PH/`1BL%E at Y"`/?0B8=Y____H9(.0@`#
MP at 7#````*P6:#D(`]]")1_;_-6H.0@#HH@(``/\U9 at Y"`.CE`@``_S5>#D(`
MZ-H"``#_->L.0@#HU0(``(,]B@]"``!T!VK_Z(\"``!H``("`&I`Z)4"``"C
MZPY"`&H`Z&4"``"C?@]"`*,;#T(`N.<'00"C#P]"`,<%+P]"`&</0@!H"P]"
M`.BO`@``,\!0_S5^#T(`4%!H``(``&@``0``:F1J9&@```P`:&</0@!H9P]"
M`%#HI0(``*."#T(`:@'_-8(/0@#HF0(``/\U@@]"`.B4`@``_S6"#T(`Z&L"
M``"C8P]"`#/`:@%04%!H,P]"`.A/`@``A<`/A;,```"A4P]"`(/@#Z-3#T(`
M at _@`#X0!`0``N0`0``"+->L.0@!15E^#QQ2+1 at BKBT8,JX$&`!```*T!1 at 2M
M`48$K<'X$(O0K<'X$(O8K8/X`'4%@\8(ZTF`;OP!<P3&1OP`@&[]`7,$QD;]
M`(!N_@%S!,9&_@!64%)3_S5C#T(`Z/8!``!>K<'X$(O0K<'X$%9J`%)0_S5C
M#T(`Z-L!``!>@\8$64D/A7G_____!5,/0@#I-/___X,]-P]"`!)T%K at S#T(`
M4%#HJ0$``.B&`0``Z17_____-6,/0@#_-8(/0@#H4@$``,<%B@]"`/_____I
M/_[__UBCA@]"`(-\)`0"=0MJ`.@[`0``,\#K!>A*`0``BPV&#T(`4<.A3P]"
M`(/@#Z-/#T(`P>`-BSWK#D(``_BY``$``.AF````P>@(HUL/0@#H60```,'H
M"*-7#T(`Z$P```#!Z`@-#P^O`(O8Z#T```#!Z`^)!XE/!-M'!-G^V at _;1P39
M_]H/VQ_;7P2#QPBA5P]"`*NA6P]"`*N+PZN#QPSBR?\%3P]"`.EW_O__N!-`
M(0#W+5\/0@`KT at 41$%,"HU\/0@##_R5D`$,`_R5H`$,`_R5L`$,`_R5P`$,`
M_R5T`$,`_R5X`$,`_R5\`$,`_R6``$,`_R6$`$,`_R6(`$,`_R6,`$,`_R60
M`$,`_R64`$,`_R68`$,`_R6<`$,`_R6@`$,`_R6D`$,`_R6H`$,`_R6P`$,`
M_R6T`$,`_R6X`$,`_R7``$,`_R7$`$,`_R7(`$,`_R7,`$,`_R70`$,`_R74
M`$,`_R78`$,`_R7<`$,`_R7@`$,`_R7D`$,`_R7H`$,`_R7P`$,`````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M`+*EK__]____^__P_P``__]'_________[__Y?]:15)/"````/_^__]%[__Q
MX$OV,MY'_K,RWF]OJY>6C-^/C9"8C9Z2WY**C(O?G9K?C8J1WXJ1FYJ-WZB6
MD<S-\O7;R%I%4D\B````K[K__[/^^O]@VBIQ6D523P(````?_W%>]/[]YO_U
M____]____________O____[____]____O_____[___W___[__________/_U
M__________G___O________]____6D523P(```#__^___^_________O____
M___[_Z________S_D_[__UI%4D\&````___Z_Q/___]:15)/%````+RPN[K_
M_____^_______O__]?____G__UI%4D\#````W___G[N^J[[______^______
M_?___?___^___UI%4D\#````O___/]&6FYZ+GO___^_______/___?___^W_
M_UI%4D\#````O___/]&:FYZ+GO___^______^____?___^O__UI%4D\#````
MO___O]&-FI.0G/___^______^O___?___^G__UI%4D\#````O___KUI%4D_0
M````?(/;]_^+LGR#V_?^B_T4L$'#_[W_J:!&RO___U,)+U4=!9?_W___E;\7
M9_?__WH_BOO,/Q310&;_O?]4^D?T__]4^D?T__]4^A?\__]4?#^;5!3T`,IF
M_[W_%Y+W__]'_O___SWS_Q>4^O__?`:;\'AU____=/S:("`@(,*ROK:SBLUT
MO/O:`"`@(,+?N:VPBMR9=+SWF=H@`)G"LL6*ZG0,=,)>_[W_=O+__[W_#%L6
M@/[__W3\VB`@("#"K;ROJXK(=+S[V@`@(`#"WZNPQ8K7%T_X__]V\OO_O?\7
M(O[__WP'__![MO[__SCZ:O^]_P`````6SO[__Q;*_O__%\'\__]\!__P>MC^
M__]TRF+_O?^94IG"___P>^K^__]\PFK_O?__\'H*____%U#[__]>?O^]_W3B
M9O^]_Q>6^___?`<`\'L/____%T_[__]'^O___T3E_[W_%[/[__]\!P#P>RS_
M__\7!_G__\+-RL_?\'H\____1_G___]$]_^]_Q?;^___?`<`\'M4____%R_Y
M___"S<K/W_!Z9/___U[__[W_=.)>_[W_%P3\__]\!P#P>WW___\76/G__\+-
MRL_?BHETRF+_O?]TZE[_O?]\%<W,-G014[[$#8BBP_^+W\/UB at UT/G0B%T/\
M__]\!P"+N!>3^?__PLW*S]^*Q!0M1_G___]$\?^]_Q=E_/__?`<`B]H7M?G_
M_\+,RLO?BN8X^FK_O?______=,)B_[W_S#]41_[___\\1[*RLK(\%VW\__]\
M!IN-EA<,_O__?`?_BJ`7>/S__UY^_[W_=.)F_[W_%[[\__]\!P"+M!=S_/__
M1_K___]$Y?^]_Q?7_/__?`<`B\T7)_K__\+-R\_?BME'^?___T3K_[W_%_?\
M__]\!P"+[1='^O__PLS+S]^*^4?^____/$>QL;&Q/)]T%)6;`,I6_[W_%P[Z
M__]Z/_![F/[__UR"_[W_09G_O?]TPE;_O?_\PH+_O?]&]/___PQ;E?^7?___
M_Y7[E?^5_Y?___\_`,I6_[W_%QKZ__^_\'O9_O__MUR*_[W_E\WK__^5OQ=C
M^O__>C_P>_[^__]<>O^]_Y7_EX;_O?^7_^O__P#*>O^]_P#*BO^]_Q>!^O__
M>C_P>QW___]^PH;_O?__Z___C<@`RHK_O?\7=OK__Y7_EW____^5_97_E?^7
M____/P#*5O^]_Q>9^O__O_![6/___[=<BO^]_Q1G=`I\.?=2="=2="]TRGK_
MO?]T,?SRAO^]_U+$/(O]%/=2Q#V+N'P1^WP1_)E2Q`Z,]IG"\O6+';$4#G0!
M=`I\.?=T\OO_O?]\%O>NJ`Q;F4?R]9E4H*:5_Y>&_[W_KJ@`RHK_O?\7'OO_
M_Q3B`,IZ_[W_%TG[__\`RHK_O?\7)/O__YY'`````#P`RGK_O?\79OO__P#*
MBO^]_Q=!^___GLP_/)]T#'0!_`9T,'P^F\PMF73AF7X<("!T$9E^!+FMB[ET
M"IE^!*RJBZQT"IE^!+&ZBX]T"IE^!+R\\'MF____=`J9?@2]O/![5O___W0*
MF7X$\O7P>T;___]T"KG$#O!\&____Q16F5)2VB`@``#"L++%WXI4%RK___\4
M;)E24MH@("`@PKVUNKR*85+:(```_\*KQ=__BFX73?___Q:2````F5)2VB`@
M("#"J*RXK8I^4MH@("`@PK"JKZSP>H\```"94IG"Q=_P>IL````7 at ____Q;(
M````F5)2V@``___"Q=____!ZJP```!>@____%N4```"94E+:(```_\*\Q=__
M\'J[````%[W___\6`@$``%+"\O7R]?!ZQ````*]'I]*LCU1'GI&,E%1'GL7?
MIE291YJ,F51\/?&G5+V]=NI^_[W_GLP_/)Y'`````#QT"G3"9O^]__P%4U6]
M?@5Y]/__B/O#]8H./'07E?^OK`#*<O^]_P#J=O^]_WP'`(KJ%Q']___"S-C_
M_XK[=#H4)$<`````/'2CV\MTL]O'/'2CV\]VXG+_O?^GI*RO?#S$=/S\)WP\
M^W;B=O^]_SR7_Z___Y6_%T3]__]Z/_![K_W__UQZ_[W_E9L`RE;_O?\79/W_
M_WH_\'O4_?__7(+_O?]!HO^]_W3"5O^]__S"@O^]_T;V____#%N5_Y=_____
ME?R5_Y7_E____W\`RE;_O?\7</W__[_P>Q7^__^W7([_O?^7V/^]_P#*CO^]
M_Q>3_?__O_![/?[__[=<V/^]_Y7_E?^5_Y7]E?\`RH[_O?\7Z_W__WH_\'M@
M_O__7.#_O?^5_Y7_E?^5^P#*X/^]_Q?M_?__>C_P>XO^__]<W/^]_T'#_[W_
M=,)Z_[W_1N;___\,6\PM7MC_O?^OF432_YD(#,PVF70W=O+4_[W_=^K0_[W_
M=,K<_[W_HOP1=.K4_[W_3[)51O#____$"HK[S3\4_E-U'S\7_=O`BOW[O_O?
M574[/Q_[V\]U!\0*BOO-/Q3^4W4?/Q?[V_#U.-O`BOW[O_O?574[/Q_]VP-U
M!\0*BOO-/Q3^4W4?/Q?YV_SU.-O`BOW[O_O?574[V\"*_?N_^]]5MHILF4?R
M]9E4M8I]7]#_O?_#_XO4^]]5S#_,+<PV7]#_O?],_`D,=3=[&XO]F;Y%_O__
M_SGZT/^]__\6J0```$&J_[W_1O?___\,6W0PU/)Z_[W_=O+'_[W_S"W,-E['
M_[W_F407_)D(#)ET-W;RS_^]_W;JR_^]_W3*>O^]_P#ZS_^]_T(7_/__?,+/
M_[W__HKY=-++_[W_=#IT(1=]`@``?`<`B_7\"@#RS_^]_XHK`,K<_[W_%UW_
M__\`RN#_O?\74/___P#*CO^]_Q=;____`,IZ_[W_%Y;___\\E?^5PP#*6O^]
M_P#*<O^]_Q?*____?`<`BQETXEK_O?]T_#R?=,IB_[W_="E^/:OT__]3Q`V,
M[L/_B at BQ=`%T#`Q;1_____]4GCP`VL/_O/\`VK__O/\`VK?_O/\`VK/_O/\`
MVJ__O/\`VJO_O/\`VJ?_O/\`VJ/_O/\`VI__O/\`VIO_O/\`VI?_O/\`VI/_
MO/\`VH__O/\`VHO_O/___UI%4D]!````K:RZJ_+UN[ZKOO+UK["LJ_+U\O71
M\O7_6D523P<````-"F)E9VEN(#8T-"!(87!P>3DY+F5X90T*8`T*96YD#0I<
M4VMA+F5X90!<;&ES=&4N<VMA`/___UI%4D]F````@__\_\/__/]:15)/`P``
M`'?__/^W__S_6D523P4```!I__S_8?_\______]/__S_.?_\_R/__/\5__S_
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MGI&;DYK_6D523R@```##__O__O____W____]____U__[_\__^__'__O_F/_^
M_[_]_O^Y__O_M/_[_____O^2GIN3D]&[L[/_DIZ6D_^1FHB,_UI%4D]L````
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MQPW'!\<!QUI%4D]&`@``14Y$(`"VC-^6B]^>WXF6C8J,T]^>WXB0C9+3WY[?
MBXV0E9Z1P-^RL*JKTK*PJJO?MX:=C9:;W]><UM^LCYZ1C)2>W\[&QL;1_Z.(
MC)"<E,S-T9N3D_^CK)2>T9N3D_^CK)2>T9J'FO^LD)F+B)Z-FJ.REIR-D(R0
MF8NCJ):1FY"(C*.\BHV-FI&+J9J-C):0D:.MBI&PD9R:_P``````````````
M````````````````````````````````````````````````````````````
M``````````````````````````!+15).14PS,BYD;&P`3&]A9$QI8G)A<GE!
M`$9R965,:6)R87)Y`$=E=%!R;V-!9&1R97-S````````````````````````
M`````````````````````````````````````T``````````````````````
M``````````````<`````````````````````````````````````````````
M````````````````````````````$;NJ_P````!(87!P>2!.97<@665A<B`Q
M.3DY("$A````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M``````````````````#X``,`9``#``````````````````4!`P"P``,`````
M````````````$@$#`,```P`````````````````=`0,`\``#````````````
M````````````````*`$#`#0!`P!&`0,`7@$#`'(!`P!^`0,`D`$#`)X!`P"L
M`0,`P@$#`-`!`P#F`0,`]@$#``P"`P`:`@,`*`(#`#0"`P!$`@,``````%`"
M`P!B`@,`=`(#``````""`@,`C@(#`*`"`P"R`@,`P@(#`,H"`P#>`@,`\`(#
M``(#`P`0`P,`(`,#```````T`P,``````$M%4DY%3#,R+F1L;`!!1%9!4$DS
M,BYD;&P`55-%4C,R+F1L;`!'1$DS,BYD;&P`````5W)I=&5&:6QE````56YM
M87!6:65W3V9&:6QE````1V5T5VEN9&]W<T1I<F5C=&]R>4$`````1V5T36]D
M=6QE2&%N9&QE00````!#;W!Y1FEL94$```!'9710<F]C061D<F5S<P````!%
M>&ET4')O8V5S<P```$=E=$9I;&53:7IE````1V5T36]D=6QE1FEL94YA;65!
M`````$QO8V%L06QL;V,`````0W)E871E1FEL94UA<'!I;F=!`````$=E=%9E
M<G-I;VY%>$$```!'9713>7-T96U$:7)E8W1O<GE!````0W)E871E1FEL94$`
M``!#;&]S94AA;F1L90```$QO8V%L1G)E90```$UA<%9I97=/9D9I;&4```!2
M96%D1FEL90````!296=3971686QU945X00````!296=#<F5A=&5+97E%>$$`
M``!296=#;&]S94ME>0```%)E;&5A<V5$0P```%)E9VES=&5R0VQA<W-!````
M`%!O<W11=6ET365S<V%G90```%!E96M-97-S86=E00````!'971$0P```$1I
M<W!A=&-H365S<V%G94$`````1&5F5VEN9&]W4')O8T$`````0W)E871E5VEN
M9&]W17A!````4VAO=U=I;F1O=P````!5<&1A=&57:6YD;W<`````5')A;G-L
M871E365S<V%G90````!39710:7AE;%8`````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M```````````````````````````````````````````````````!`&@!```6
M,$8P7C!P,'LPD#"B,*LPL3"W,,8PS##6,-PP`3$',0PQ$C$8,38Q2#%/,54Q
M6S%A,7,Q>3%_,8PQDC&8,:LQL3'-,=TQXC'P,04R$C(=,BTR.S)-,EHR;#*;
M,J4RKC*X,L8R\C(.,RXS-C-#,THS4#-9,X`SAC.2,Y at SM3/5,^0S\#/U,_LS
M!C0;-"HT-C0[-$$T3#19-&4T=S1\-((TE#29-)\TL32V-+PTSC34--HTMC7!
M-<DUTC77->(U[S7\-0<V$C8=-B at V9#8S-DXV6C9?-FDV;S9S-G at VAC:@-J4V
ML#:X-L,VSC;8-N0V]C;^-A(W;3>(-Y\WJC>R-\DWSS?:-^DW!C at -.!4X'C at R
M.#\X=CA\.(LXFSBG.*XXM#BZ.,`XQCC,.-(XV#C>..0XZCCP./8X_#@".0 at Y
M#CD4.1HY(#DF.2PY,CDX.3XY1#E*.5`Y5CE<.6(Y:#EN.0``````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
M````````````````````````````````````````````````````````````
*````````````````
`
end




From siona at chiark.greenend.org.uk  Wed Nov 24 09:43:24 1999
From: siona at chiark.greenend.org.uk (Sion Arrowsmith)
Date: 24 Nov 1999 14:43:24 +0000 (GMT)
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <383979F7.25D6@cs.bham.ac.uk>
Message-ID: <SxA*J5Vdo@news.chiark.greenend.org.uk>

In article <383979F7.25D6 at cs.bham.ac.uk>,
Stuart Reynolds  <S.I.Reynolds at cs.bham.ac.uk> wrote:
>I'd like there to be do..while loops. Only using while loops can make
>for some-more-ugly-than-necessary code (lots of people complain about
>than from time to time).

I'm quite happy to be stuck with just while loops (rarely using
'do ...' loops in languages which support them anyway) but the
standard 'while 1: ... break' idiom is not aestheticly pleasing.

>Oh yeah! I'd I disklike writing 'self.' eveywhere.

I use 'I' which has the added bonus of making references to
the self object stand out when you haven't hacked your Python
mode to highlight 'self'.

-- 
\S -- siona at chiark.greenend.org.uk -- http://www.chaos.org.uk/~sion/
  ___  |  "Frankly I have no feelings towards penguins one way or the other"
  \X/  |    -- Arthur C. Clarke
   her nu become? se bera eadward ofdun hl?ddre heafdes b?ce bump bump bump



From ivnowa at hvision.nl  Mon Nov 29 15:37:35 1999
From: ivnowa at hvision.nl (Hans Nowak)
Date: Mon, 29 Nov 1999 21:37:35 +0100
Subject: __getslice__
In-Reply-To: <3842C0EB.6357@cs.bham.ac.uk>
Message-ID: <199911292036.VAA14482@axil.hvision.nl>

On 29 Nov 99, Stuart Reynolds wrote:

> >>> def square(x):
>         return x*x
> 
> >>> xvals = [0,1,2,3,4,5,6,7]
> >>> a = UnaryFnSequence(square, xvals)
> >>> xvals[1:]
> [1, 2, 3, 4, 5, 6, 7]
> >>> a
> ( (0,0), (1,1), (2,4), (3,9), (4,16), (5,25), (6,36), (7,49) )
> >>> a[1:]
> ( (0,0), (1,1), (2,4), (3,9), (4,16), (5,25), (6,36) )
> 
> Note that the first item is still there. The last one's gone instead.

Did you try UnaryFnSequence(square, [2,3,4])? You'll get a surprising 
result:

((0,0), (1,1), (2,4))

>     def __repr__(self):
>         s='( '
>         for x in range( len(self) ):
>             s = s+'('+`x`+','+`self._fn(x)`+'), '
>         return s[:-2] + ' )'

This is the culprit... You take a list of N elements, and then you 
loop over *a range 0..N-1*... rather than the list! Try replacing the

  for x in range(len(self)):

with

  for x in self._inputs:

This might do what you intended. ^_^

Hope this helps,

--Hans Nowak (zephyrfalcon at hvision.nl)
Homepage: http://fly.to/zephyrfalcon
You call me a masterless man. You are wrong. I am my own master.



From da at ski.org  Wed Nov  3 23:58:22 1999
From: da at ski.org (David Ascher)
Date: Wed, 3 Nov 1999 20:58:22 -0800 (Pacific Standard Time)
Subject: __str__ vs. __repr__
In-Reply-To: <199911040157.UAA11908@eric.cnri.reston.va.us>
Message-ID: <Pine.WNT.4.05.9911032048030.109-100000@david.ski.org>

On Wed, 3 Nov 1999, Guido van Rossum wrote:

> Any suggestions?

I certainly appreciate the justification for the current behavior, which I
didn't really understand either.

I have two suggestions which would make the current read-eval-print
behavior somewhat more palatable to me:

 -- container types (lists, dicts, tuples) should have both repr and str
    methods, even if they differ only in the methods of the contents that
    they call.  You can make the str methods use the repr-type display
    (with brackets/parens/braces and commas), as that wouldn't break
    anyone's expectations (since that's what is currently being used), and
    would 'solve' the current inconsistency.

    I understand str output is meant to be human readable -- that
    doesn't mean that []'s and ,'s can't be used.

 -- make the str of a long integer not have the postfix L.  After all, if
    the __str__ is supposed to be for human consumption, the L is no more
    useful than the quotes around strings.  

The issue of determining the printability of characters is one *I*
wouldn't touch with a 10-foot snake.

--david






From aa8vb at yahoo.com  Wed Nov 24 09:38:36 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Wed, 24 Nov 1999 09:38:36 -0500
Subject: Set Title Bar Icon with Tkinter?  (was: Tk title bar)
In-Reply-To: <80v0d8$agu$1@nnrp1.deja.com>
References: <80v0d8$agu$1@nnrp1.deja.com>
Message-ID: <19991124093836.B1426604@vislab.epa.gov>

hhh14021978 at my-deja.com:
 |I'm new to python-tk and I love it. Moreover, I want to change the icon
 |in the title bar. Is there any way to do that?

What windowing system?  X/Windows (UNIX)?  or MSWin?

If X, this is a function of the window manager.  You need to configure your
window manager to set this icon based on the application's class name
and/or window title.

If MSWin, sorry, you're on your own.

Randall



From exeq at my-deja.com  Wed Nov  3 22:18:30 1999
From: exeq at my-deja.com (exeq at my-deja.com)
Date: Thu, 04 Nov 1999 03:18:30 GMT
Subject: Tkinter and Solaris 7 Sparc [Q]
References: <7vqblc$ham$1@nnrp1.deja.com> <Pine.SO4.4.02.9911031716390.17271-100000@bulldog5.d.umn.edu>
Message-ID: <7vqtu5$ugu$1@nnrp1.deja.com>

In article
<Pine.SO4.4.02.9911031716390.17271-100000 at bulldog5.d.umn.edu>,
  Aaron J Reichow <reic0024 at bulldog5.d.umn.edu> wrote:
>
> > I just get an error saying that libX11.so.6.1 is needed. Is there a
way
> > to build tkinter without X11? Or I just need to download and install
> > X11?
>
> Don't you already have X11 installed, with Solaris (that would be that
> which drives CDE/Solaris Desktop/OpenView)?  And if the answer is no,
and
> you're running simply from the console, why would you need Tkinter,
> which provides a graphical interface?
>
> Aaron
>
>

Hi Aaron.

I have installed the entire Solaris distribution, including CDE,
Openwindows et al. :-) But it isn't X11 6.1 as the tk lib needs.
I am getting the following error message:

          ../libpython1.5.a -L/usr/local/lib -ltk8.0 -ltcl8.0
-L/usr/openwin/li
ld: warning: file libX11.so.6.1: required by /usr/local/lib/libtk8.0.so,
not fod
Undefined                       first referenced
 symbol                             in file
Py_Main                             python.o
ld: fatal: Symbol referencing errors. No output written to python
*** Error code 1
make: Fatal error: Command failed for target `link'
Current working directory /usr/local/src/Python-1.5.2/Modules
*** Error code 1
make: Fatal error: Command failed for target `python'


Thanks for your answer.

/B


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



From pinard at iro.umontreal.ca  Tue Nov 30 10:45:36 1999
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 30 Nov 1999 10:45:36 -0500
Subject: automatic "lotjes trekken"
In-Reply-To: "Tim Peters"'s message of "Tue, 30 Nov 1999 01:59:48 -0500"
References: <000401bf3b00$7ddf7ae0$c92d153f@tim>
Message-ID: <oq1z98aspr.fsf@titan.progiciels-bpi.ca>

"Tim Peters" <tim_one at email.msn.com> writes:

> [...] with a belly full of far too much turkey.

Strange.  I was imagining you rather slim.  Are you? :-)

> where p[:] ranges over all permutations of range(N) [...]

By the way, I do not remember any standard library helper to enumerate all
permutations (or combinations, or arrangements) of a given list (or tuple).
Ideally one at a time, of course.  Would this be useful?

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





From philh at vision25.demon.co.uk  Sat Nov 13 08:31:19 1999
From: philh at vision25.demon.co.uk (Phil Hunt)
Date: Sat, 13 Nov 99 13:31:19 GMT
Subject: __str__ vs. __repr__
References: <000001bf275c$02370d20$1d2d153f@tim> <7vvni7$19q0$1@nntp6.u.washington.edu> <382BEC79.EC65913D@maxtal.com.au> <382C6430.8DC07E3E@lemburg.com> <01fd01bf2dce$d2c03cf0$f29b12c2@secret.pythonware.com>
Message-ID: <942499879snz@vision25.demon.co.uk>

In article <01fd01bf2dce$d2c03cf0$f29b12c2 at secret.pythonware.com>
           fredrik at pythonware.com "Fredrik Lundh" writes:
> 2) what's the reason byte code cannot be used to recover
> the original source code

Perhaps because two different (but obviously functionally
identical and very similar) source codes could produce the same
byte code.

-- 
Phil Hunt - - - - - - - - -  philh at vision25.demon.co.uk
Moore's Law: processor speed doubles every 18 months
Gates' Law: software speed halves every 18 months




From rparker at bserv.com  Fri Nov  5 16:09:31 1999
From: rparker at bserv.com (Robert Parker)
Date: Fri, 05 Nov 1999 21:09:31 GMT
Subject: Printing from Python to Windows Printers
Message-ID: <fIHU3.641$f66.886@198.235.216.4>

I am trying to find out how I can print to a windows printer from a Python
application.  Has anyone used any of the Com modules to gain access to the
printer api in windows?  Are there any modules that have been written that
would help save time when we want to send font information only to the
printer driver not have to write in the escape sequences and print directly
to the parallel port?







From akuchlin at mems-exchange.org  Sun Nov 14 14:40:11 1999
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: 14 Nov 1999 14:40:11 -0500
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net>
Message-ID: <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us>

usenet-mm-yyyy at julesallen.com (Jules Allen) writes:
> One of the main problems I've had with understanding Python's depth
> is finding information and modules. 

Here's a bit of history, which may not be known to people who've
recently joined the Python community.

The issue of indexing modules is an old one, and I'm still astounded
that we've never converged on a solution, though everyone always
complains about this.  Originally you'd upload things to
ftp.python.org (and occasionally people still do this), or, if your
project had a Web page of its own, you could send a note to the
maintainers of python.org, and they'd add a link to
http://www.python.org/download/Contributed.html .

This centralized system has been problematic for a long time; the
python.org maintainers are all quite busy, & their time is better
spent hacking on Python code.  A decentralized system would obviously
scale better, and the Locator-SIG was started in 1995.  Looking
through the SIG archive is probably useful:

ftp://ftp.python.org/pub/python-unmirrored/maillist-archives/locator-sig/locator-sig.archive

Nothing was ever really completed; Paul Everitt's prototype never
became widely used.  The backbone of Trove
(http://www.tuxedo.org/~esr/trove/) is partially implemented, but
there's no user interface yet, ESR is too busy to work on it much,
and no one else is really helping with it.

IMHO this isn't a technical problem; it's not too hard to write the
scripts something like freshmeat.net or CPAN.  The problem is getting
acceptance from the community; people need to start using it to look
for things, and that acts as a strong incentive for people to enter
their new modules into it.  For example, in the Perl community it's
deeply ingrained that you always, *always* release things on CPAN, and
follow CPAN's naming scheme.  We just need to push the Python
community very hard to enter their data.

http://www.vex.net/~x/parnassus/ is *very* nice, and it might be the
only solution required.  Apparently a good deal of work has been put
into seeding the initial database, and that's important, because it
makes the index immediately useful.  Yet I hadn't heard about it until
now; is the site still beta?  Did I miss a c.l.py.announce article
about it?  

There are various issues with it -- I don't know how scalable the
underlying code is, the graphic design could be simpler, vex.net's
reliability and bandwidth are unknown, and we can argue about the
classification hierarchy *forever* -- but it's still valuable.

We need to just pick a solution and stick with it; Parnassus is as
good as any, and has the incalculable advatange of being implemented.
I'd like to encourage everyone who maintains Python code, or writes
Python tutorials and docs, to go to http://www.vex.net/~x/parnassus/ .
Is everything you've done in the index?  If not, add the missing
items.  Can we prominently link to it from python.org, so people begin
to use it?

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
I am the Head and I demand your shrunken experiences now!
    -- In ENIGMA #1: "The Lizard, The Head, The Enigma"



From b2blink at hotmail.com  Fri Nov 26 10:58:21 1999
From: b2blink at hotmail.com (Ulf Engstrøm)
Date: Fri, 26 Nov 1999 10:58:21 CET
Subject: SMTP
Message-ID: <19991126095821.87186.qmail@hotmail.com>

I believe I got the To: thingy working with using the simple mail at mail.ma 
(Name Of Person)
At least it works on our server, but since I'm behind a firewall I can't 
check it on an outside server.
Guess I'll have to get a shell with Pythonaccess soon...:)
Best regards

Ulf Engstr?m

http://www.b2b-link.com
ulf.engstrom at b2b-link.com

______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com



From donn at u.washington.edu  Tue Nov 23 17:26:09 1999
From: donn at u.washington.edu (Donn Cave)
Date: 23 Nov 1999 22:26:09 GMT
Subject: Method objects question
References: <slrn83lvr7.e7.bmcd@ripley.sulaco.net> <slrn83ku3j.2je.bmcd@ripley.sulaco.net> <slrn83krsu.2i0.bmcd@ripley.sulaco.net> <383a7309@news.uni-bielefeld.de> <81eha1$rf6$1@nntp6.u.washington.edu>
Message-ID: <81f4a1$o62$1@nntp6.u.washington.edu>

Quoth bmcd at es.co.nz (Ben Caradoc-Davies):
| On 23 Nov 1999 17:01:53 GMT, Donn Cave <donn at u.washington.edu> wrote:
|> It doesn't seem to matter where a function is defined, only the scope
|> where it's found.  The normal and most convenient way to get a function
|> into the class name space is certainly to define it there, but any other
|> way will serve as well - it doesn't matter if you define it outside the
|> class, or inside.
|> How could we justify anything different, in a dynamic "run time" system
|> like Python?
|
| Sure, the class behavious is excellent. But the point is, why should functions
| found in a class namespace be converted into methods, but functions found in an
| instance namespace not converted?
|
| For example, suppose you have a running application containing two instances of
| the same class. You need to patch your application on the fly, performing
| *different* modifications on each object. You send a signal to your
| application, and when it is next idle it execfiles your patch script.
|
| This script can modify each instance's variables, rebinding them to entirely
| new objects. In particular, it is possible to hide class variables by adding
| instance variables of the same name. However, with the current behaviour, it is
| not possible to bind instance variables to function objects and override class
| variables with the same name (bound to function objects) without changing any
| other code, because the rest of the program will expect these instance
| atributes to be *method objects*, not ordinary function objects. It is only
| possible to change the class variable, and thus change the behaviour of *all*
| instances of that class.

Well, it's possible, though perhaps not ideal if this were a cornerstone
of your design.  You just have to add the function first to the class
namespace, by whatever name is convenient;  then, you can add another
reference to it in the instance by the name you want, hiding the original
class attribute.  As long as the assignment that creates the second
reference locates the function through the instance, you get a bound
method.

e.g., given class X and instance x,
  x.bindme = x.bindme_new_7       # bound to instance x
  x.bindme = X.bindme_new_7       # unbound

So you're forced to use that class name space, but it's accessible
and if you're worried about collisions, some system can presumably
be invented to avoid them if you're desperate enough.
   
| In a nutshell, I don't see any fundamental reason why the
| search-instance-namespace then search-class-namespace attribute resolution
| order couldn't be used to allow method overriding on an instance-by-instance
| basis. With the current, function-to-method conversion rules, this is
| impossible.
|
| Actually, I'm more curious about why this design decision was made. (Or did it
| just happen that way!)

I sure don't know how anything gets decided.  But I'm pretty sure I've
used this the other way - that is, where the newly added function really
is just a function, not a method.  I don't think that's terribly exotic,
and it's possible due to the way this works.

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu



From damien_morton at my-deja.com  Sun Nov 14 17:23:39 1999
From: damien_morton at my-deja.com (damien_morton at my-deja.com)
Date: Sun, 14 Nov 1999 22:23:39 GMT
Subject: NT admin scripts with python
References: <000201bf2e34$26d93750$020a10ac@p350.gnaa.net> <80lhgl$22gn$1@hub.org>
Message-ID: <80ncp9$rnm$1@nnrp1.deja.com>

For doing NT admin scripts, I have found that System Scripting Host
from www.netal.com is extremely usefull.

You can happily program it in Python, VBscript, Perl or whatever. Its
basically a framework for running scripts as NT services.

In article <80lhgl$22gn$1 at hub.org>,
  z at NO-MAIL.net wrote:
> On Sat, 13 Nov 1999 16:06:50 -0800, geoff nordli <geoff at gnaa.net>
wrote:
> >Does anyone know if there is a reservoir of python scripts on the
net.  Kind
> >of like Perl's CPAN
>
> It's not like CPAN but you might try http://www.vex.net/~x/parnassus/
>
> ...
>
>


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



From gusdvg at usa.net  Sun Nov 14 03:52:03 1999
From: gusdvg at usa.net (Gusdvg)
Date: Sun, 14 Nov 1999 08:52:03 GMT
Subject: itoa on python?
Message-ID: <382e77d5.26188609@news.remarq.com>

How can I change an integer (or any other type of number) to a string?
In C I used itoa but I couldn't find it in the Python documentation.

Thanks




From wware-nospam at world.std.com  Mon Nov 29 18:45:42 1999
From: wware-nospam at world.std.com (Will Ware)
Date: Mon, 29 Nov 1999 23:45:42 GMT
Subject: Python complaints
References: <1268241794-29392927@hypernet.com>
Message-ID: <FLzGo6.AAn@world.std.com>

Gordon McMillan (gmcm at hypernet.com) wrote:
: Now that you've had some coffee this morning, do you see 
: any problems with this, Will?

Uh, yeah, never mind. There is perhaps a last vestige of utility in
the idea. I dimly recall once seeing a number theory text that, over
several hundred pages, formally defined addition of integers from
more fundamental principles. The reasoning leaned heavily on functions
for the successor and predecessor of an integer. Conceivably there
might be some value in adding successor and predecessor functions to
immutable types where they had obvious meanings. Not that I ever do
number theory myself, nor am I aware of anybody using Python for it,
so this isn't a wish list thing, this is just idle conjecture.
-- 
 - - - - - - - - - - - - - - - - - - - - - - - -
Resistance is futile. Capacitance is efficacious.
Will Ware	email:    wware @ world.std.com



From bellman at lysator.liu.se  Mon Nov 29 12:20:38 1999
From: bellman at lysator.liu.se (Thomas Bellman)
Date: 29 Nov 1999 17:20:38 GMT
Subject: cgi.py inefficiency (with fix)
References: <80u861$ngs$1@newsy.ifm.liu.se> <81989f$1m42$1@hub.org>
Message-ID: <81ucl6$7tj$1@newsy.ifm.liu.se>

T. Moddletin at news.vex.net (T. Moddletin) writes:

> On 17 Nov 1999 12:48:01 GMT, Thomas Bellman <bellman at lysator.liu.se> wrote:
>> searches to find things.  Is there any reason why this is done,
>> instead of having a dictionary of lists?

> After some thought i suddenly realised why this may be so. In fact
> it is possible that a CGI may recieve mulitiple values for the 
> same key. In this circumstance obviously a python dictionary will
> not function, as if you assign to a key it will overwrite the
> previous contents. However, i just offer this as a theoretical 
> reason, as i've not looked at the code or how it is handled.

That is why I suggested a dictionary of *lists*.

> Of course whether this would be worth the trouble in most cases
> I don't know. Perhaps i shouldn't spout off without having 
> read the code to actually know what it all does. Nor 
> have I read your diff to see what it actually does. (-:

No, you shouldn't, because my patch *does* handle that case.

-- 
Thomas Bellman,   Lysator Computer Club,   Link?ping University,  Sweden
"I refuse to have a battle of wits with an   !  bellman @ lysator.liu.se
 unarmed person."                            !  Make Love -- Nicht Wahr!



From greg.ewing at compaq.com  Tue Nov  2 08:11:46 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Wed, 03 Nov 1999 02:11:46 +1300
Subject: Changing cursors...
References: <1270748767-19272684@hypernet.com>
Message-ID: <381EE312.EAA128AE@compaq.com>

Gordon McMillan wrote:
> 
> Xtian Muirhead writes:
> >
> > I've been having a play with Tkinter, and I've ported across a
> > program that I wrote while learning VB, but I've had one problem.
> > There's one step that has a chunk of processing which takes a few
> > seconds to complete. In the old program, I changed the cursor to
> > an hourglass at the start, and reset when the program finished.
> > Trying to do the same in the new version doesn't seem to work:

I managed to kludge my way around this problem using the
following technique:

    root = Tkinter._default_root
    x, y = root.winfo_pointerxy()
    w = root.winfo_containing(x, y)
    if w:
        old_cursor = w['cursor']
        w.configure(cursor = "watch")
        w.update()
	w.configure(cursor = old_cursor)

This finds the widget that the mouse pointer happens to be
over, changes its cursor, and forces this to become the
current cursor by doing an update. The change back again
doesn't take effect until the event loop is returned to
or another update occcurs.

(This works on Windows, and will probably work on the Mac
(if Tkinter works there at all). It almost certainly won't 
work on unix under X, because X maintains the cursor 
continuously.)

Greg



From a.eyre at optichrome.com  Wed Nov  3 09:41:13 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Wed, 3 Nov 1999 14:41:13 -0000
Subject: Please Help!!!
In-Reply-To: <7vnsjg$jih$1@nslave1.tin.it>
Message-ID: <002801bf2609$7a0f7de0$3acbd9c2@peridot.optichrome.com>

> I'm building a simple telnet client...
> I don't know how to catch the key pressed without read it from stdin...
> I read the manual about the Event class but i can't understand a lot...
> Can you help me, maybe with an example?

What platform?

Text based/GUI based?

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From mlh at vier.idi.ntnu.no  Fri Nov 26 09:02:02 1999
From: mlh at vier.idi.ntnu.no (Magnus L. Hetland)
Date: 26 Nov 1999 15:02:02 +0100
Subject: decompress a bunch of .gz files
References: <81jc8i$ilp$1@nnrp1.deja.com> <000501bf374e$116bafb0$3acbd9c2@peridot.optichrome.com> <14397.18620.87516.687622@gargle.gargle.HOWL>
Message-ID: <y0jn1s1xsf9.fsf@vier.idi.ntnu.no>

Charles G Waldman <cgw at fnal.gov> writes:

> Adrian Eyre writes:
>  > > I'm searching for a way to decompress all .gz
>  > > files in a directory automatically.
>  > 
>  > Unix:
>  > 
>  > #!/bin/sh
>  > for file in *.gz; do gunzip $file; done
> 
> Or just:
>    
>  gunzip *.gz

Or just

   gunzip *

:)

--

  Magnus          Echelon jamming noise:
  Lie             FBI CIA NSA Handgun Assault Bomb Drug Terrorism
  Hetland         Special Forces Delta Force AK47 Hillary Clinton 



From robin at jessikat.demon.co.uk  Thu Nov  4 17:18:27 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Thu, 4 Nov 1999 22:18:27 +0000
Subject: embedding stubbed tcl
References: <bLd7vSA1euH4EwDS@jessikat.demon.co.uk>
 <8ZCT3.18814$1v5.202282@newscontent-01.sprint.ca>
 <VVo0SGAJExH4EwyM@jessikat.demon.co.uk> <3821E345.FC6C372F@chello.nl>
Message-ID: <Ik5jOCAzYgI4Ewad@jessikat.demon.co.uk>

In article <3821E345.FC6C372F at chello.nl>, Jan Nijtmans
<j.nijtmans at chello.nl> writes
>Robin Becker wrote:
>
>> I thought stubs were for extensions. The tkinter case is slightly the
>> wrong way round as it starts tcl up.
>
>Stubs can very well be used for applications as well. For example,
>I already used it to Stubify the Tcl plugin, so I don't need to
>recompile it when Tcl8.3 comes out ;-).  The only problem is that
>Tcl doesn't yet contain a portable way to use dlopen()/LoadLibrary()
>or whatever this function might be called.
>
ok so what is the minimal way to do this?
-- 
Robin Becker



From ScherBi at BAM.com  Thu Nov  4 06:55:56 1999
From: ScherBi at BAM.com (ScherBi at BAM.com)
Date: Thu, 4 Nov 1999 06:55:56 -0500 
Subject: Extending Python dynamically
Message-ID: <3F8BA119B48AD11197470008C7A493CB98B636@HQOBGEX1>

> I have downloaded DCOracle extension. I have RH6.0, and (at 
> least until now) I have not installed the source distribution of Python. I
have
> installed only the binary RPMs. I unpacked DCOracle in a directory and
> managed to build the so files (Buffer.so and oci_.so).
> How to install then this DCOracle package? I have been trying 
> to figure out but no success.


Create a directory somewhere in your pythonpath (import sys; print
sys.path), call it DCOracle.
In this directory put the two .so files, and all the .py files that came in
the DCOracle package.

Then you may import DCOracle.

Let me know if you need more help.

Bill Scherer



From mhammond at skippinet.com.au  Tue Nov  9 05:05:11 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Tue, 09 Nov 1999 10:05:11 GMT
Subject: SWIG for Win95/Python1.5.2
References: <808n8t$gv1$1@nnrp1.deja.com>
Message-ID: <rlSV3.32675$pf4.29226@news-server.bigpond.net.au>

Add a "/D __WIN32__" to the C compiler, and it should fix it.

At the top of the file there should be a series of #ifdef's etc that will
end up with:

# define SWIGEXPORT(a,b) __declspec(dllexport) a b

This magic will solve the problem.  In a nutshell, this is exposing the
initfoo function as a public function from the DLL.

Mark.

moonseeker at my-deja.com wrote in message <808n8t$gv1$1 at nnrp1.deja.com>...
>Hi!
>
>I tried to compile the example (renamed as test1) in the SWIG introduction:
...
>It compiled and linked correct, but if I try to import it in Python 1.5.2:
>
>>>> import test1module
>Traceback (innermost last):
>  File "<stdin>", line 1, in ?
>ImportError: dynamic module does not define init function (inittest1module)






From kanzow at gmd.de  Fri Nov 26 07:33:05 1999
From: kanzow at gmd.de (Peter K.)
Date: Fri, 26 Nov 1999 13:33:05 +0100
Subject: Running apps with Zope 2?
References: <80v5g8$5a6$1@mulga.cs.mu.OZ.AU> <3834101A.FF39F64C@infercor.no> <8E83A8056PaCmAnRDLM@194.2.0.33> <slrn83jits.cj8.quinn@seniti.ugcs.caltech.edu> <383E7D7A.1B48C7F1@gmd.de>
Message-ID: <383E7E01.2EF074CF@gmd.de>

"Peter K." wrote:

> Sorry, there was a mistake in the code:

> File m.py:
>
> class Application:
>     "Application"
>     def hello(self):
>         "Hello"
>         return "Hello World from my Zope Object"
>
> app=Appl()

is wrong, it should be
app=Application()

Thanks, Peter





From rvollmert at gmx.net  Tue Nov 16 15:47:04 1999
From: rvollmert at gmx.net (Robert Vollmert)
Date: Tue, 16 Nov 1999 21:47:04 +0100
Subject: problems with __getattr__ & __setattr__
In-Reply-To: <3F8BA119B48AD11197470008C7A493CB98B67D@HQOBGEX1>; from ScherBi@bam.com on Tue, Nov 16, 1999 at 02:24:08PM -0500
References: <3F8BA119B48AD11197470008C7A493CB98B67D@HQOBGEX1>
Message-ID: <19991116214703.C7366@krikkit.vollmert.home>

Hi,

On Tue, Nov 16, 1999 at 02:24:08PM -0500, ScherBi at bam.com wrote:
> I am having trouble using __getattr__ and __setattr__ in some classes.  I've
> boiled it down to the code below.
> It dumps core under 1.5.2 on Linux or WinNT.
> 
> I think maybe it's looping, if so, how does one go about doing this? 
> (I'm assuming it's clear enough what I'm trying to do.)
> 
> If I comment out the __gettattr__ operation, it doesn't dump core.  Instead
> I get an AttributeError: eggs raised at the 'self.eggs[key] = value' call in
> __setattr__.  This is what leads me to belive there's something circular
> going on here.

in __init__, 'self.eggs = {}' calls __setattr__, where the assignment
to self.eggs[key] calls __getattr__(self, 'eggs'), as self.eggs
doesn't exist yet. This is also why you get an attribute error if
__getattr__ is commented out.

In __getattr__, the reference to self.eggs causes __getattr__ to be
called again. This continues until infinity (well, almost ;-). 

To fix this, you would probably best use self.__dict__['eggs'] instead
of self.eggs, as this skips __getattr__/__setattr__ calls.

Hope this is correct and helps,
Robert

> -------------------------------------------------------------
> 
> class spam:
> 
>     def __init__(self):
>         self.eggs = {}
>     
>     def __getattr__(self, key):
>         try:
>             return self.eggs[key]
>         except:
>             raise
>         
>     def __setattr__(self, key, value):
>         try:
>             self.eggs[key] = value
>             return 0
>         except:
>             raise
> 
> 
> 
> if __name__ == '__main__':
> 
>     s = spam()
>     print 'instance of spam created.' 
> 
> ------------------------------------------------------------------


-- 
Robert Vollmert                                      rvollmert at gmx.net



From mikael at isy.liu.se  Wed Nov 24 10:28:43 1999
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Wed, 24 Nov 1999 16:28:43 +0100 (MET)
Subject: Python complaints
In-Reply-To: <1268673769-3409210@hypernet.com>
Message-ID: <XFMail.991124162843.mikael@isy.liu.se>

On 24-Nov-99 Gordon McMillan wrote:
 > >>> def p(score):
 >  ...  print "You scored %d point%s" % (score, score==1 and ' ' 
 >  or 's')
 >  ...
 > >>> p(1)
 >  You scored 1 point
 > >>> p(3)
 >  You scored 3 points

Cool, but I would never had understood that code without testing it. 
However, cool is not necessarily the same as good.

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    24-Nov-99
Time:    16:26:03

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



From framiere at netcom-service.com  Mon Nov 15 08:15:02 1999
From: framiere at netcom-service.com (Florent Ramière)
Date: Mon, 15 Nov 1999 14:15:02 +0100
Subject: Python ISAPI
References: <80oq15$ln8$1@oceanite.cybercable.fr> <ftSX3.4566$Xo.24188@news-server.bigpond.net.au>
Message-ID: <80p15v$nff$1@oceanite.cybercable.fr>

Hello,

    would it be a good thing to make python ISAPI compliant :) ?
    In fact i am using cgi under NT/IIS, and it takes looooooooots of times
to reimport
    every time and every times the same objects ...
    Any idea to make the imported classes persistant ?

Florent.

Mark Hammond <mhammond at skippinet.com.au> wrote in message
news:ftSX3.4566$Xo.24188 at news-server.bigpond.net.au...
> AFAIK there is no ISAPI module - but Python is an AXScript language, so it
> works fine in ASP (hence no real demand for ISAPI)
>
> Mark.
>
> Florent Rami?re wrote in message <80oq15$ln8$1 at oceanite.cybercable.fr>...
> >Hello,
> >
> >    i am looking for an python ISAPI, but could not find any.
> >    Could you tell me where i can get such a goodie ?
> >    Does anybody ever worked on this ?
> >
> >Thanks.
> >
> >
>
>





From darrell at dorb.com  Mon Nov 15 09:38:27 1999
From: darrell at dorb.com (Darrell)
Date: Mon, 15 Nov 1999 09:38:27 -0500
Subject: Binary search
References: <38300986.E734E9E2@obop.com.pl>
Message-ID: <092c01bf2f77$1481ae40$0100a8c0@rochester.rr.com>

Check out bisect.py

--Darrell
----- Original Message -----
From: Przemys?aw G. Gawro?ski <gawron at obop.com.pl>
Subject: Binary search


> I read somewhere that Python has a build in binary search.
> How can I use it ?





From fdrake at acm.org  Mon Nov 29 17:02:38 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Mon, 29 Nov 1999 17:02:38 -0500 (EST)
Subject: Python bugs
In-Reply-To: <51C9F3C392D0D111A15600A0C99E64A9AF569C@localhost>
References: <51C9F3C392D0D111A15600A0C99E64A9AF569C@localhost>
Message-ID: <14402.63486.483442.646147@weyr.cnri.reston.va.us>

Skip said:
 > I believe os.environ is a class to allow constructs like
 > 
 >     os.environ["FOO"] = "PYTHON"
 > 
 > to work like
 > 
 >     os.putenv("FOO", "PYTHON")

Bjorn Pettersen writes:
 > Which is a good thing, but shouldn't preclude it from being used as a
 > dictionary in the update statement (although the fix probably have to be to
 > Dict.update rahter than os.environ).

  Very much so.  os.environ does exactly the right thing; {}.update is 
simply too simple.  ;-)  Shouldn't be had to fix, given a little
motivation.  The implementation should retain the special case of
"real" dictionaries, but also implement a generic version that uses
the mapping API.
  I'm sure Guido would be glad to receive a patch.  ;-)  ;-)


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From gerrit.holl at pobox.com  Wed Nov  3 09:30:35 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Wed, 3 Nov 1999 15:30:35 +0100
Subject: Please Help!!!
In-Reply-To: <7vnsjg$jih$1@nslave1.tin.it>
References: <7vnsjg$jih$1@nslave1.tin.it>
Message-ID: <19991103153035.A17161@optiplex.palga.uucp>

Francesco wrote:
> I'm learning Python...
> I'm building a simple telnet client...
> I don't know how to catch the key pressed without read it from stdin...
> I read the manual about the Event class but i can't understand a lot...
> Can you help me, maybe with an example?

Use the curses module, read the curses HOWTO.
    
regards,
Gerrit.
-- 
linuxgames.nl.linux.org	 All about games on Linux in Dutch (under construction).
www.nl.linux.org	       The Dutch resource for Dutch Linux documentation.
www.nl.linux.org/discoverb		    Learn foreign words and definitions.
www.nl.linux.org/~gerrit/asperger	 I'm an Asperger, go here for more info.
www.nl.linux.org/~gerrit			    About me and pictures of me.



From dorland at lords.com  Wed Nov  3 20:16:15 1999
From: dorland at lords.com (Eric Dorland)
Date: Thu, 04 Nov 1999 01:16:15 GMT
Subject: Exceptions from callbacks
References: <381CA8E5.5D69E64@lords.com> <7virj4$b7b$1@towncrier.cc.monash.edu.au> <381FAE6D.73F426FE@lords.com> <19991103071727.A87523@vislab.epa.gov>
Message-ID: <3820DE5B.A5500330@lords.com>

Randall Hopper wrote:
> I've done this, and I got the impression that Jonathan understood your
> question.  What he described works and it's what I'm using.
> 
> To extend your description, you don't say "Wake up, Python!" from the
> callback.  You say it from the C function(s) from which you entered the C
> library that invoked the callback.  Return NULL, and the exception is
> restarted with the Python stack frames saved in the callback in-tow.
> 
> To make this concrete, for a C library API that might potentially invoke a
> callback (which in-turn would relay to Python) would look like this:
> 
>   (Python) -> HandlePendingMessages -> MYLIB -> Callback -> (Python)
> 
> And your wrapper code might look like this:
> 
> ------------------------------------------------------------------------------
>   PyObject *_wrap_MYLIBHandlePendingMessages( PyObject *self, PyObject *args )
>   {
>     PyErr_Clear();
> 
>     <<< call C function here >>>
> 
>     /*  If a Python callback threw an exception, restart its propagation  */
>     /*    in the controlling Python.                                      */
>     if ( PyErr_Occurred() )
>       return NULL;
> 
>     ...
>   }
> 
> ------------------------------------------------------------------------------
>   static void _wrap_MYLIBCallback( void )
>     /*  C callback used to relay to Python callable objects.  */
>   {
>     ...
> 
>     /*  Pass the buck onto Python  */
>     pyresult = PyEval_CallObject( pyfunc, pyargs );     /*  Call Python     */
>     Py_DECREF( pyargs );                                /*  Trash arg list  */
>     if ( pyresult )
>       Py_DECREF( pyresult );                            /*  Trash any result*/
> 
>     /*  If the callback threw an exception, it will be propagated across    */
>     /*    the C layer ( we check PyErr_Occurred() on the other side).       */
>   }
> ------------------------------------------------------------------------------
> 
> If the Python callback tossed an exception, then it's still lying around
> when _wrap_MYLIBCallback() returns.  We then return through the C library,
> which eventually returns to _wrap_MYLIBHandlePendingMessages, who
> recognizes that a Python exception is pending and restarts its propagation
> in the controlling Python stack frames.
> 
> Note that if you have a C library layer that isn't Python-exception-aware,
> then it could go on to perform useless or invalid work after a Python
> callback throws an exception.  In this case you'll need to "make" it aware.
> That is, have it call PyErr_Occurred(), or turn Python exceptions into C
> library errors inside of your callback wrappers.
> 
> Hope this clarifies things a bit.

I understand what you mean exactly. The problem is I can't do it that
way. Basically, the way the library works is that you register a bunch
of callbacks and then call a function (eg. Loop()) that doesn't return.
The HandlePendingEvents (its equivalent I mean :)) isn't exposed by the
library, so I can't wrap it :). That's way I want a way to force the
interpreter to realize an exception is there. Is there a way, or a work
around, to do this?

-- 

Eric Dorland
dorland at lords.com



From python-list at teleo.net  Tue Nov  2 03:59:43 1999
From: python-list at teleo.net (Patrick Phalen)
Date: Tue, 2 Nov 1999 00:59:43 -0800
Subject: commas vs. spaces. vs. the legions of new lines
In-Reply-To: <007f01bf250c$72dd3c40$f29b12c2@secret.pythonware.com>
References: <1270647588-25358408@hypernet.com> <007f01bf250c$72dd3c40$f29b12c2@secret.pythonware.com>
Message-ID: <99110201025204.01264@quadra.teleo.net>

[Fredrik Lundh, on Tue, 02 Nov 1999]:

::     "Welcome to the Unix Reconstruction Project. Our
::     goal is quite simply to reimplement the classic Unix
::     command set in pure Perl, and to have as much fun
::     as we can doing so."

This is quite possibly the scariest idea I have ever heard.



From ivanlan at callware.com  Sat Nov  6 12:40:10 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Sat, 06 Nov 1999 10:40:10 -0700
Subject: making the Python case to my dept.
References: <9911051740130D.02530@william.jinsa.org> <3823FC43.CF7FBD71@python.net> <99110609085702.02970@quadra.teleo.net>
Message-ID: <382467FA.577E70A0@callware.com>

Hi All--

Patrick Phalen wrote:
> 
> [Thomas A. Bryan, on Sat, 06 Nov 1999]:
> :: William Ahern wrote:
> ::
> :: > Could anybody suggest some substantive info to make the case (or not make it).
> :: > Also, the department will need an IDE (Win95/Mac), so if anybody knows where to
> :: > find one....
> 
> [snip well-reasoned, even-handed answer]
> 
> Since this is a FAQ, I'd vote for preserving Thomas' answer somewhere
> on the python.org Web site, so people could point to the URL when the
> question comes up again.
> 

I agree.  His answer is nearly perfect.  Thanks, Thomas.

Take care,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From ivanlan at callware.com  Mon Nov 15 15:34:29 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Mon, 15 Nov 1999 13:34:29 -0700
Subject: Line Benchmarks Tkinter vs. wxPython
References: <01bf2fa7$02992ca0$8a42c3d1@Cyberus.cyberus.ca>
Message-ID: <38306E55.62A6D135@callware.com>

Hi All--

Gordon Williams wrote:
> 
> Hello All,
> 

[snip]

> 3)      wxPython beats Tk by a very! large margin even when Tk is called
> directly.  Direct drawing onto to screen is roughly 3 times faster than
> using a bitmap.
> 
> Thanks to:
> Greg Landrum for providing some information on direct calls to tk.
> Ionel Simionescu for providing the wxPython code which I was able to
> modify.
> 

Could you post the benchmark code?  Or provide us with a link where it
can be downloaded?

Thanks,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours
----------------------------------------------



From dieter at handshake.de  Sun Nov 14 13:07:10 1999
From: dieter at handshake.de (Dieter Maurer)
Date: Sun, 14 Nov 99 18:07:10 GMT
Subject: TreeWidget 0.01
Message-ID: <mt2.0-4346-942602830@news.informatik.uni-bonn.de>

TreeWidget 0.01 displays (dynamically changable) tree structures
as interactive hierarchical content tables.
(i.e. the display is similar to the left part of the MS Windows Explorer
and other hierarchical tree browsers.)


TreeWidget is based on Greg McFarlane's excellent
Python MegaWidget Framework (URL:http://www.dscpl.com.au/pmw/).
It is similar to the TreeExplorer in Doug Hellmann's
PmwContribD package
(URL:http://members.home.net/doughellmann/PmwContribD/index.html),
but supports dynamic modification of the displayed tree.


More information and download at:

	URL:http://www.handshake.de/~dieter/pyprojects/treewidget.html


- Dieter

<P><A HREF="http://www.handshake.de/~dieter/pyprojects/treewidget.html">TreeWidget
0.01</A> - displays PMW-based hierarchical content trees; similar to
TreeExplorer from the PmwContribD package.  (14-Nov-99)

-- 
----------- comp.lang.python.announce (moderated) ----------
Article Submission Address:  python-announce at python.org
Python Language Home Page:   http://www.python.org/
Python Quick Help Index:     http://www.python.org/Help.html
------------------------------------------------------------



From gerrit.holl at pobox.com  Mon Nov 29 11:46:31 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Mon, 29 Nov 1999 17:46:31 +0100
Subject: IDLE & Linux
In-Reply-To: <uQh04.4688$tI2.97781@ord-read.news.verio.net>; from rkaiser@wwa.com on Sun, Nov 28, 1999 at 04:18:32PM -0600
References: <B466ECEC.F8E%Dan@Grassi.com> <uQh04.4688$tI2.97781@ord-read.news.verio.net>
Message-ID: <19991129174631.B4550@stopcontact.palga.uucp>

Ralph Kaiser wrote:
> 
> Dan Grassi <Dan at Grassi.com> wrote in message
> news:B466ECEC.F8E%Dan at Grassi.com...
> > I am trying to get IDLE to run.  The docs say that it is automatically
> > installed but that is just not true.  There is a /usr/bin/idle but when it
> > runs it quits complaining that it can't find
> > /usr/lib/python1.5/site-packages/idle/idle.py
> 
> Hi,
> 
> /usr/bin/idle is just a one line script to run idle.py Use "find .-name
> idle.py"
> to locate it. on my system it's
> /usr/local/src/Python-1.5.2/Tools/idle/idle.py.

find .? Are you mad!
locate idle

Or just have a look:
/usr/src/Python
/usr/doc/python*
/usr/doc/packages/python-1.5.2 # suse

Oh, sorry, I'm assuming you're running Linux... are you?

regards,
Gerrit.

-- 
"If a machine couldn't run a free operating system, we got rid of it."

  -- Richard Stallman (Open Sources, 1999 O'Reilly and Associates)
  5:45pm  up  2:29, 16 users,  load average: 1.52, 1.36, 1.08



From jesse at student.usyd.edu.au  Sun Nov  7 22:50:33 1999
From: jesse at student.usyd.edu.au (Jesse Sweeney)
Date: Mon, 08 Nov 1999 03:50:33 GMT
Subject: IDLE and rlcompleter [Was: Re: making the Python case to my dept.]
References: <9911051740130D.02530@william.jinsa.org> <3823FC43.CF7FBD71@python.net> <m3puxno6g0.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <382523fc.14124721@news.usyd.edu.au>

On 06 Nov 1999 17:50:23 +0000, Michael Hudson <mwh21 at cam.ac.uk> wrote:

[ snip ]

>``import rlcompleter''!!
>
>It makes life *so* much easier, and is the reason I still use the
>basic interpreter rather than idle.
>
>Hmm, implementing completion in idle? It has ``expand word'' but
>that's not what I want. Would this be easy? Pointers appreciated.

Coincidentally, I had the same idea just a few days ago, and after a bit of
tinkering, I came up with two different IDLE extensions. I've put the files on
the web at

http://www-personal.usyd.edu.au/~jsweeney/python .

The first (RLComplete.py) attempts to mimic the behaviour of readline and
rlcompleter in the Python shell, and the second (AutoComplete.py) is a bastard
child of rlcompleter and the current "Expand Word".

Obviously, unlike "Expand Word" these only work in the Python shell, not in the
file-editing windows. (Now that I think about it, this would probably be pretty
useful in those editing windows, but that would take some partial compilation of
some sort I suppose, which is way too much work for me.)

To add these to IDLE, put the two files in your IDLE directory, (or just the one
you want -- I'm flexible!), and add "RLComplete" and "AutoComplete" to the list
of names in extend.py (also in your IDLE directory).

Problems and inconsistencies are documented in the module docstrings. Let me
know of any bugs.

	Cheers, Jesse.



From tbryan at python.net  Wed Nov 24 06:28:10 1999
From: tbryan at python.net (Thomas A. Bryan)
Date: Wed, 24 Nov 1999 06:28:10 -0500
Subject: Upgrading Python on Windows
Message-ID: <383BCBCA.5804D43E@python.net>

I have Python-1.5.1 installed on a Windows95 machine.

I now want to upgrade to Python-1.5.2.  I can't find any 
documentation about upgrading on a Windows box.  I don't need 
the Python-1.5.2 installation anymore.  I've already downloaded 
py152.exe, but I'd like advice before I proceed.

Should I uninstall Python-1.5.1 first?  How do I do that (run 
unwise.exe?)?  Should I also uninstall the current version of 
Tcl/Tk and reinstall it with the Python-1.5.2 installation?

I-can-troubleshoot-a-UNIX-install,-but-Windows-scares-me-ly yours

---Tom



From bparsia at email.unc.edu  Sat Nov 27 12:27:54 1999
From: bparsia at email.unc.edu (Bijan Parsia)
Date: Sat, 27 Nov 1999 12:27:54 -0500
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <867lj85g4c.fsf@g.local> <383CC835.B585EBDD@Lugoj.Com> <m3ogcjrn41.fsf@atrus.jesus.cam.ac.uk> <383D9C63.3A7E5806@Lugoj.Com> <86ogci2tfm.fsf@g.local> <al8g0xt5k8f.fsf@sirppi.helsinki.fi> <86aeo0283v.fsf@g.local>
Message-ID: <1e1xf0n.w5dt9t1qn1y4qN%bparsia@email.unc.edu>

Gareth McCaughan <Gareth.McCaughan at pobox.com> wrote:

> Markus Stenberg wrote:
> 
> [I said:]
> >> Garbage collectors have got a *lot* better since the days when
> >> GC meant half-second pauses waiting for the machine to sweep
> >> through all its memory. Let alone 20 seconds.
> > 
> > Where can I admire such a beast? At least none of the Java
> > implementations(*) I have tried(**) have one. say, for 128MB of trash, it's
> > consistent N second sweep (0.5-10, depending on hardware).
> 
> In a modern Lisp implementation. Modern Smalltalks are probably
> just as good. (Unfortunately, "modern" might have to imply
> "commercial" here; I'm not sure of the quality of the GC in
> the various free implementations of those languages.)

Yep, they're as good. As for trying 'em out, the are non-commercial
(i.e., limited to non-commercial uses unless you purchace a licence) of
VisualWorks (http://www.cincom.com) for many platforms (the latest
version, 5i, is available for Linux and Windows, but not yet for
PowerMac). VW allows you to do all sorts of things with it's gc via a
class called, IIRC, Memory Policy.

There is a non-commercial version of Dolphin Smalltalk
(http://www.object-arts.com/), albeit for Windows only. But it's a nice
smallish Smalltalk (VW can be overwhelmning, particularly at first).
(I'll add that Dolphin is rather inexpensive and is definitely worth it)
I believe Smalltalk MT may have a demo version.

As for Free/Open Source, there's Squeak (http://www.squeak.org), which
is massively cross platform (even runs on WinCE & Sharp's ICruise).
There's a very clear description of Squeak's gc in the "Back to The
Future" paper
(ftp://st.cs.uiuc.edu/Smalltalk/Squeak/docs/OOPSLA.Squeak.html). An
edited quote:

"Two policy refinements reduced the incremental garbage collection
pauses to the point where Squeak became usable for real-time
applications such as music and animation...We have been using this
storage manager in support of real-time graphics and music for over a
year now with extremely satisfactory results. In our experience, 10
milliseconds is an important threshold for latency in interactive
systems, because most of the other critical functions such as mouse
polling, sound buffer output and display refresh take place at a
commensurate rate."

There also was a Real Time Squeak project which was, alas, canceled, but
the bits and pieces are filtering their way into the main distribution.
A paper describing the work was given at the OOPSLA '99 "VM Workshop"
(http://www.squeak.org/oopsla99_vmworkshop/). The particular paper is
"Experience Report: Building a Real Time OS Kernel in Smalltalk". There
have been other RT Smalltalks, I believe.

Gnu Smalltalk has be revived, but I don't *believe* it's settled on a gc
mechanism yet, although it may just borrow Squeak's (in concept if not
in code).

Note: I was confused by the term "non-deterministic" wrt to GC. My first
thought was that it referred to the fact that, generally, one doesn't
know exactly *when* garbage will be collected, only that it *will* be,
at some point, if necessary ;) This is certainly an important issue for
such matters as finalization. But it seems to have been used to refer to
the "pausing" problem. But I don't think that's necessarily
"non-deterministic", per se. It isn't that hard to predict how long a
stopping gc will take for a given memory space! ;)

Cheers,
Bijan Parsia



From sk at fyemf.fys.ntnu.no  Mon Nov  8 10:01:31 1999
From: sk at fyemf.fys.ntnu.no (Steinar Knutsen)
Date: 8 Nov 1999 15:01:31 -0000
Subject: What are Aspects? (Re: How can I make my assertions smarter?)
References: <Pine.WNT.4.05.9911032042440.109-100000@david.ski.org> <7vvq63$ehm$1@nntp4.atl.mindspring.net> <3826C32A.FB47EF6@compaq.com> <y0jpuxlnicz.fsf@vier.idi.ntnu.no>
Message-ID: <806okb$qkr$1@fyemf.fys.ntnu.no>

In article <y0jpuxlnicz.fsf at vier.idi.ntnu.no>, Magnus L. Hetland
<mlh at vier.idi.ntnu.no> wrote:

>Greg Ewing <greg.ewing at compaq.com> writes:
>
>> Smacks of the legendary come-from statement, somehow...
>
>Sounds cool... Only a myth?

Well, it depends. INTERCAL defines COME FROM. :)
http://www.tuxedo.org/~esr/intercal/

(For those who haven't heard of it, INTERCAL is perhaps the most twisted
programming language ever implemented.)
-- 
Steinar



From hniksic at iskon.hr  Tue Nov 16 04:44:05 1999
From: hniksic at iskon.hr (Hrvoje Niksic)
Date: 16 Nov 1999 10:44:05 +0100
Subject: list.without()?
References: <y0jln7zi9pb.fsf@vier.idi.ntnu.no>
Message-ID: <9t9g0y6ojoa.fsf@mraz.iskon.hr>

mlh at vier.idi.ntnu.no (Magnus L. Hetland) writes:

> Is there a function or method in the standard language that can
> non-destructively return a list without a specified element?

What's wrong with destructively removing the element from a copy?  It
will be faster than using `filter' and friends as other proposed, and
guaranteed to work.  In fact, (shallow) copying is the pretty much
standard way of turning any destructive operation non-destructive.

If you want it bundled in a function, you can do this:

def without(list, element):
    copy = self[:]
    copy.remove(element)
    return copy

If you want to use classes, you can inherit from UserList.

If your only problem with this approach is having a temporary
variable, then you don't need to lobby for adding a new method -- you
can simply lobby for the `remove()' method to return self.  In that
case, you could write:

    newlist = somelist[:].remove(element)

Or, for that matter:

    newlist = somelist[:].sort()

Or:

    for elem in dict.keys().sort():
        ...

But neither is likely to be successful.  Adding a new builtin method
for a task that can be accomplished with two lines of code doesn't
sound promising.  And the destructive methods rarely return self, "to
remind you [that they modify the list in place]."



From olivier at scripps.edu  Wed Nov 10 15:37:32 1999
From: olivier at scripps.edu (Olivier Roche)
Date: Wed, 10 Nov 1999 12:37:32 -0800
Subject: random function
Message-ID: <3829D78B.7FC142F4@scripps.edu>

Hi everybody

By looking at the index of the python documentation index i haven't seen
any
random function description. Could somebody tell me if a random function

to generate random number (between 0 and 1) exists ?

Thanks,

Olivier




From henk_p at my-deja.com  Tue Nov 16 13:48:07 1999
From: henk_p at my-deja.com (henk_p at my-deja.com)
Date: Tue, 16 Nov 1999 18:48:07 GMT
Subject: PyApache
References: <Pine.LNX.4.20.9911151623540.28300-100000@fep132.fep.ru>
Message-ID: <80s8sk$cr7$1@nnrp1.deja.com>

Maybe you can try my Python Script Interface,
It provides persistent python scripting to any CGI capable webserver.

It was announced recently on the python announce news group.

You can download it directly from:

http://huizen.dds.nl/~d50/psi0.02.tgz

Let me know if it is usable to you.

Henk.


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



From brainsite at altern.org  Thu Nov  4 08:42:53 1999
From: brainsite at altern.org (brainsite at altern.org)
Date: 4 Nov 1999 13:42:53 GMT
Subject: FREE Porn, without membership!!! No pop-ups!!! Everythings FREE!!
Message-ID: <7vs2gt$28jm2528@michener.tea.state.tx.us>

***********************************
* XXXMonkey - 100% Free Pure Porn *
***********************************

Goto XXXMonkey and enjoy the taste of FREE porn. There is no membership, just pics you freely can save on YOUR harddisk. From TEENS to ASIANS, everything is there.

Go and check:
http://get.to/muchxxx (with 1 small pop-up)
http://watch.at/xxxmonkey (no pop-ups)

Keep in mind: This site is just new!! It must grow, so the archive isn't big YET!!




From wtanksle at hawking.armored.net  Sat Nov 27 15:04:56 1999
From: wtanksle at hawking.armored.net (William Tanksley)
Date: 27 Nov 1999 20:04:56 GMT
Subject: Python complaints
References: <000601bf36be$f1bf5e80$292d153f@tim> <81ivhe$12l$1@mach.vub.ac.be>
Message-ID: <slrn840egc.bt2.wtanksle@hawking.armored.net>

On 25 Nov 1999 09:29:18 GMT, Thomas Hamelryck wrote:
>Tim Peters (tim_one at email.msn.com) wrote:
>: There is one (& only one) area where Python's indentation gets in my way a
>: bit:  conceptual blocks that can't be written as blocks.  These typically

>Right. Another example occurs in OpenGL, eg.

>glBegin(GL_POLYGON)
>   glVertex2f(-0.5, -0.5)
>   glVertex2f(-0.5, 0.5)
>   glVertex2f(0.5, 0.5)
>   glVertex2f(0.5, -0.5)
>glEnd()

Good point -- the forced indentation here would be annoying.  A
try/finally block wouldn't be totally amiss, but perhaps a function
definition would be better.  Either way is merely a workaround, but the
try/finally block has some useful meaning: it communicates that the stuff
in the block is in some way atomic.

>This makes the code more readable, but it's not allowed in Python. I've 
>got used to the indentation rule in Python, and I'm not annoyed by it that
>much anymore, but I still think it is a major mistake.

I think it's the best decision that could have been made.

All hail Guido.

;-).

>Thomas Hamelryck//Free University Brussels (VUB)

-- 
-William "Billy" Tanksley



From kaleissin at nvg.org  Sat Nov 27 22:42:52 1999
From: kaleissin at nvg.org (Kaleissin)
Date: 28 Nov 1999 03:42:52 GMT
Subject: Anyone looked at pike?
Message-ID: <slrn84193k.6mc.kaleissin@rhiann.nvg.ntnu.no>

Specifically: why they claim it's faster?
http://pike.idonex.se/manual/tutorial/tut_introduction.html#introduction.3

Though, why anyone would want Yet Another C-Like Syntax is beyond me...<wink>


tal.



From kvijayan at calum.csclub.uwaterloo.ca  Sat Nov 27 01:08:13 1999
From: kvijayan at calum.csclub.uwaterloo.ca (Kannan Vijayan)
Date: 27 Nov 1999 06:08:13 GMT
Subject: [].extend only accepts list?
References: <slrn83uhd2.6lc.quinn@groat.ugcs.caltech.edu>
Message-ID: <81nsgd$mk5$1@watserv3.uwaterloo.ca>

Quinn Dunkan (quinn at groat.ugcs.caltech.edu) wrote:
> Is there any reason why the extend method of lists shouldn't accept any
> sequence?

the extend method checks to see if the argument is a list.  You
can easily get past this by using the list() function

y = ["one", 2, "three"]

y.extend(list(("four", 5.0, 3*2)))

y -> ["one", 2, "three", "four", 5.0, 6]

you can use this method with any sequence, including strings, and tuples.

-kannan



From dgrisby at uk.research.att.com  Fri Nov  5 12:01:49 1999
From: dgrisby at uk.research.att.com (Duncan Grisby)
Date: 5 Nov 1999 17:01:49 -0000
Subject: C++ extension modules on NT
References: <7vuqa8$fvi$1@pineapple.uk.research.att.com>
Message-ID: <7vv2ht$hd3$1@pineapple.uk.research.att.com>

In article <7vuqa8$fvi$1 at pineapple.uk.research.att.com>,
 Duncan Grisby  <dgrisby at uk.research.att.com> wrote:

[...global constructors in DLL...]

>On NT, however, the constructors are never called, and the global
>objects contain garbage. The document "Extending and Embedding the
>Python Interpreter" mentions this problem in section 1.11, and says
>that the problem doesn't occur if the Python interpreter's main()
>function is compiled with C++. Unfortunately, this appears to be
>incorrect, and the constructors still aren't called.

Following up to my own post, it turns out that the problem I was
having was something else (to do with dllimport), and the constructors
_are_ being called. So now my question is: are there any platforms for
which the comment in the documentation is true?

Cheers,

Duncan.

-- 
 -- Duncan Grisby  \  Research Engineer  --
  -- AT&T Laboratories Cambridge          --
   -- http://www.uk.research.att.com/~dpg1 --



From pycloutier at videotron.ca  Mon Nov 29 00:56:01 1999
From: pycloutier at videotron.ca (Pierre-Yves Cloutier)
Date: Mon, 29 Nov 1999 00:56:01 -0500
Subject: Recursion?
Message-ID: <38421570.9C3D9AEC@videotron.ca>

Can I use recursion in python?

thanks

Pierre-Yves Cloutier




From kyriacou at umbc.edu  Tue Nov 23 15:39:30 1999
From: kyriacou at umbc.edu (Stelios Kyriacou)
Date: 23 Nov 1999 15:39:30 -0500
Subject: whrandom seed question
Message-ID: <81eu22$411al$1@umbc7.umbc.edu>

Hi, 

I tried today the random number generator in python 
and have the following question: 

Shouldn't the following 3 lines produce the same number? 

>>> whrandom.seed(0,0,0); print whrandom.random() 
0.39613674715 
>>> whrandom.seed(0,0,0); print whrandom.random() 
0.170188890936 
>>> whrandom.seed(0,0,0); print whrandom.random() 
0.831213988577 
>>> 
 

I am used to the matlab randn function which does produce the 
same number since i restart the seed: 
>> rand('seed',0); rand(1)
ans =
    0.2190
>> rand('seed',0); rand(1)
ans =
    0.2190
>> 
 

Am I missing something??? 

Thanks very much in advance. 

Stelios 
  




From tim_one at email.msn.com  Wed Nov 24 17:09:25 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 24 Nov 1999 17:09:25 -0500
Subject: CRC-module
In-Reply-To: <383C1089.1D0F7C18@bibsyst.no>
Message-ID: <000801bf36c8$92ee19a0$292d153f@tim>

[Thomas Weholt]
> Ok, so I`ve looked into zlib.crc32 and zlib.adler32. They seem easy
> enough to use, but I thought crc-codes had characters and numbers in
> them, not just a plain integer like the methods above return.

Bits is bits.  Whether you choose to view them as integers, or characters,
or little stick men struggling to avoid falling into little holes is all in
your head.  That is, don't worry about it!

> ( As you can see, I`m a complete ass on this subject,

In that case, I'll be toilet paper.

> but don`t have time to do the proper research myself, and was hoping
> for a "quick fix" ... )
>
> A friend of mine mentioned that I should try SHA-1 instead, for more
> accuracy. Can anybody give me an example on how to compute crc-codes,
> using zlib or preferrably some more accurate method, for single files ??
>
> If this is all it takes :
>
> crc = module_name.crc_method(file)
>
> and comparison is done like :
>
> if (crc1 == crc2): print "Equal."
> else : print "Different."
>
> then all I need is the name of the most effective/accurate module to
> use.

That's about it.  Strictly it should be:

if crc1 == crc2:
    print "Probably equal."
else:
    print "Definitely different."

Equality of hash codes does not guarantee equality of files under any
method, it merely makes it likely.  Throw two random but unequal files at
one of the xxx32 methods, and there's (exactly) 1 chance in 2**32 (about 1
in 4 billion) that they'll return the same hash code.  The odds for MD5 are
about 1 in 2**128, and for SHA-1 about 1 in 2**160.

Like everything else, it's a tradeoff.  CRCs are cheap to compute, but very
easy to fool on purpose (it's trivial to construct any number of distinct
files that have the same CRC hashcode).  The other methods are expensive to
compute, but extremely difficult to fool on purpose (in fact, it's believed
to be computationally infeasible).  This reflects their design goals:  CRCs
were designed to catch transmission errors (accidents of nature), while MD5
and SHA-1 were designed to stop malicious people with large piles of money
(accidents of nature <wink>).

> If, for some strange reason, I should use one module instead of another,
> that info would be interesting too.

In practice, I like a two-step approach:

1. Compute CRCs.  If they're different, the files are definitely different,
   and you get out cheap.

2. Compute SHA-1.  If they're different too, the files are also definitely
   different.  If they're the same, then since the CRC and SHA-1 methods
   are unrelated, it's reasonable to expect that the odds of a false match
   are about 1 in 2**32 * 2**160 == 2**192 now.  This is even smaller than
   the chance Python2 will have curly-brace block delimeters.

which-in-turn-is-smaller-than-the-chance-you'll-wake-up-in-a-womb-ly y'rs
    - tim





From stalnaker at acm.org  Thu Nov 25 17:31:14 1999
From: stalnaker at acm.org (Max M. Stalnaker)
Date: Thu, 25 Nov 1999 14:31:14 -0800
Subject: urllib.urlopen-bug?
References: <943390670.1482784211@news.teleweb.at>
Message-ID: <HMi%3.46841$YI2.2342241@typ11.nn.bcandid.com>

I think you get html on a 404, so it is not really a *server* error.  But
consider:

errcode, errmsg, headers = h.getreply()
  if errcode!=200:
   return

"Markus Krutz" <krutz at cg.tuwien.ac.at> wrote in message
news:943390670.1482784211 at news.teleweb.at...
> Hi,
> I've got a problem using the urlopen-function from the urllib.
> According to the library reference, an IOError-exception should be raised
on
> two conditions:
> 1) The connection can't be made.
> 2) The server returns an error.
> On 1) an exceptions is raised, but if I do a
> urllib.urlopen("http://www.python.org/nonexistingdocument.html")
> no exception is raised, although I think this is a typical 404-Error.
> In my opinion this should result in raising an exception...
> Any hints on other ways how to check for the existence of a net resource?
> (I'm using python v1.5.2)
>
> greetings
> Markus
>





From fredrik at pythonware.com  Tue Nov  2 03:29:56 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 2 Nov 1999 09:29:56 +0100
Subject: commas vs. spaces. vs. the legions of new lines
References: <1270647588-25358408@hypernet.com>
Message-ID: <007f01bf250c$72dd3c40$f29b12c2@secret.pythonware.com>

Gordon McMillan <gmcm at hypernet.com> wrote:
> > So the unix old-timers <wink> know a billion and one ways to
> > convert Dos to Unix text files.  Why discourage someone from
> > making a pure python implementation... 
> 
> I will make the observation that one can classify Pythonistas 
> into those who think recoding common utilities in pure Python 
> is evil, and those who think it a worthy goal, with very few 
> people in the middle.

also see:
http://language.perl.com/ppt/index.html

    "Welcome to the Unix Reconstruction Project. Our
    goal is quite simply to reimplement the classic Unix
    command set in pure Perl, and to have as much fun
    as we can doing so."

</F>




From akuchlin at mems-exchange.org  Wed Nov 10 13:27:15 1999
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: 10 Nov 1999 13:27:15 -0500
Subject: statics everywhere.  Why?
References: <80cam4$phe$1@rtpnews.raleigh.ibm.com>
Message-ID: <3d9046p5h8.fsf@amarok.cnri.reston.va.us>

"Arinte" <shouldbe at message.com> writes:
> Embedding python in a c application, why do I need to make the functions
> that python call and variables for callbacks static?

In the modules that come with Python, all of these functions are
defined as static.  I think this is simply done to keep those symbols
private, to avoid causing collisions when you link libpython.a into
your application; you'd be unhappy if a get_group_num() function in
the re module conflicted with a get_group_num() function in your code.
For your own internal application, it's up to you whether such symbols
should be kept private or not.

(Note that this applies to functions and global variables defined as
'static'; the 'static' keyword means something different inside a C {
} block.)

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
The Kappamaki, a whaling research ship, was currently researching the
question: How many whales can you catch in one week?
    -- Terry Pratchett & Neil Gaiman, _Good Omens_





From salvadorej at my-deja.com  Thu Nov  4 20:12:18 1999
From: salvadorej at my-deja.com (salvadorej at my-deja.com)
Date: Fri, 05 Nov 1999 01:12:18 GMT
Subject: Problems Building Python under Cygwin
Message-ID: <7vtath$n1j$1@nnrp1.deja.com>

I'm having some problems building Python under Cygwin on Windows NT 4.

1) the first problem I had was relatively easy to deal with. Towards
the end of the make process, there's a command to copy the python
executable from the Modules directory. The problem is, the makefile
expects to copy something called 'python', but the executable produced
was 'python.exe' (presumably because I'm under Windows). So cp can't
complete this command, and the makefile chokes. Fortunately, at this
point, the make process is pretty much done, so all I had to do was go
into the Modules directory, and copy python.exe to the directory above,
and it worked. SO: did I mess up? Or is this something wrong with the
makefile/configure setup that needs to be fixed?

2) the second problem I haven't solved. What I really wanted Python for
was to build Zope from scratch. When I went to build Zope, it informed
me that I needed not the "vanilla" version of Python, but Python
compiled with thread support. OK. I go back, make clean, and try
configure --with-threads. And partway through the make, I get this:

 cd Modules ; make OPT="-g -O2" VERSION="1.5" \
                prefix="/usr/local" exec_prefix="/usr/local" all
 gcc  -g -O2 -I./../Include -I.. -DHAVE_CONFIG_H -c ./threadmodule.c
 ./threadmodule.c:38: #error "Error!  The rest of Python is not compiled
 with thread support."
 ./threadmodule.c:39: #error "Rerun configure, adding a --with-thread
option."
 ./threadmodule.c:40: #error "Then run `make clean' followed by `make'."
 make[1]: *** [threadmodule.o] Error 1
 make: *** [Modules] Error 2

Hmmmm. I thought I did that. So I re-did the configure, did the make
clean again, and then make. Same result. For good measure, I rm -rf'd
the whole Python directory, and then untarred it again, did configure
with both the --with-thread and --with-threads options, and tried to
make again. Same result.

Hints?

===================================================================
"The best laid plans of mice and men are about equal"
Email: iowa_s8ong at hot8mail.com
Remove Eights from email to reach me.


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



From mhammond at skippinet.com.au  Thu Nov 18 06:54:07 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 18 Nov 1999 11:54:07 GMT
Subject: Python in Windows CE?
References: <19991118102301.910.qmail@hotmail.com>
Message-ID: <zNRY3.509$VA6.2556@news-server.bigpond.net.au>

Ulf Engstr?m wrote in message <19991118102301.910.qmail at hotmail.com>...
>>You will probably want to check out the Python CE
>>mailing list:
>>
>>    http://www.egroups.com/group/python-ce/

>While both CE and Python attempt to be portable, it appears that the
current
>situation is not at all portable.  It appears there is no way we can write
a
>win32 Python program, and have it work on both CE, and Win95/NT.

"No way" isnt quite correct.  The Python shell also runs on the desktop.
Its true that you basically cant take a program written for the desktop and
expect it to run unchanged.  Unicode issues at least will prevent this
before Python 1.6.

>At the most basic level, simple win32 wrappers are in differently named
>modules.  For example, the CE port has win32reg for the registry functions,
>but in the existing win32 ports, they exist in win32api.  This is not to
>suggest my scheme is anywhere near ideal, but IMO a common strategy is
>needed.

Agreed, and it is slowly happening.  Many modules are built from the same
code base - indeed, all "new" modules are.  win32api doesnt exist on CE,
and, again, is unlikely to until at least Python 1.6.  But many modules are
as identical as the platform allows.

Pythonwin is not there, and wont be until at least CE3.0 :-)

>Has any solutions to this been found? And are there any other compability
>problems?

A number.  But not alot more than there are between the platforms
themselves - eg, there is _no_ language that allows you to take an existing
win32 app and have it run on CE.  C++ definately cant.  VB is probably the
closest, but I have never used it on CE.

And I have to be honest - if you need to do anything non trivial, it is
likely you will need to have a C compiler for CE, and add a few new "swig"
declarations, particularly for the GUI stuff.  At least for now :-)

Mark.






From mehta at mama.indstate.edu  Thu Nov 11 11:08:49 1999
From: mehta at mama.indstate.edu (mehta at mama.indstate.edu)
Date: Thu, 11 Nov 1999 16:08:49 GMT
Subject: docs
Message-ID: <80epm7$vbo$1@nnrp1.deja.com>

Hello,

I have noticed for the local docs of python in IE5 on winnt pages
randomly donot get rendered but are left as raw html in browser as I go
back and forth using browser buttons.  has any one experienced same
problem?


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



From wware-nospam at world.std.com  Mon Nov  8 17:54:30 1999
From: wware-nospam at world.std.com (Will Ware)
Date: Mon, 8 Nov 1999 22:54:30 GMT
Subject: Stackless python progress?
References: <38274626.5D8C1EC2@ptway.com>
Message-ID: <FKwIAv.IB1@world.std.com>

Brian Haskin (haskin at ptway.com) wrote:
> also has there been any work on combining
> 'microthreading' and the stackless stuff?

I was the one doing the microthread stuff. I've been too busy to
play with it recently. Implementing microthreads in terms of
Stackless Python is the Right Thing To Do(tm) from an engineering
standpoint, but I haven't had time to devote to it lately.
-- 
 - - - - - - - - - - - - - - - - - - - - - - - -
Resistance is futile. Capacitance is efficacious.
Will Ware	email:    wware @ world.std.com



From I.Clarke at strs.co.uk  Thu Nov 11 07:09:29 1999
From: I.Clarke at strs.co.uk (Ian Clarke)
Date: Thu, 11 Nov 1999 12:09:29 +0000
Subject: Tk Vs GTK - a comparison anyone?
Message-ID: <382AB1F9.85F78EAD@strs.co.uk>

Could anyone who is familar with using both Tk and GTK from Python give
a quick rundown of how they compare (speed, development time (ie. which
is more developer friendly?), level of integration with Python,
flexibility etc).

Cheers,

Ian.


--
Ian Clarke                        Email: I.Clarke at strs.co.uk
Homepage: http://www.gnu.demon.co.uk/
Also see: http://freenet.on.openprojects.net/
"All we see and all we seem is but a dream within a dream"



From ivanlan at callware.com  Tue Nov 30 09:48:57 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Tue, 30 Nov 1999 07:48:57 -0700
Subject: bug in tkinter canvas event handling?
References: <38429BB0.1BFFB54B@sage.att.com> <3843C0A0.94990E64@home.com>
Message-ID: <3843E3D9.9C68E2A1@callware.com>

Hi All--

Doug Hellmann wrote:
> 
> "Garrett G. Hodgson" wrote:
> >
> > ok, i'm back with another tk problem.
> > what i want to do is attach a callback to a shape in a canvas
> > that will tell me which domain level object was picked.
> >
> > seems simple, but i don't get what i expect.  in particular,
> > if i press a mouse button while pointing to one object, then drag
> > to another and release, the release event callback tells me i
> > was pointing to the first object.  i've distilled it down to a
> > small test case, attached below.
> >
> > this seems like a bug to me, but experience has shown that it's
> > probably a misconception on my part about how the event handling
> > is supposed to work.
> 
> This behavior surprised me at first, too.  It's as though the first
> canvas object is grabbing the pointer so that all events go to it.  I
> got around the problem by looking for the x,y coords of the release
> event and then finding the object(s) under that point on the canvas.
> 

Actually, this isn't a bug.  All windowing systems (the ones I know
about, anyway) behave this way by design; the down-click does indeed
grab the mouse pointer, and does not let go until the up-click (or is
that up-chuck?).  You can see this behaviour even in windows, which
hacks away at expected pointer motion with gleeful abandon.  Click on a
scrollbar, begin moving it, move the pointer out of the scrollbar,
*watch the scrollbar snap to its previous position*, move the pointer
back into the scrollbar, and finally *watch the scrollbar snap back
under the pointer.*  This is repulsive and incorrect behaviour <0.9
barf>, but it couldn't be done without a mouse pointer grab by the
object underneath the pointer on the down-click.

<m$-doing-the-wrong-thing-for-the-right-reasons>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours
----------------------------------------------



From news at dorb.com  Wed Nov 10 23:14:38 1999
From: news at dorb.com (Darrell)
Date: Wed, 10 Nov 1999 23:14:38 -0500
Subject: ASCII delimited files
References: <942231502.1914033961@news.interbusiness.it>
Message-ID: <cnrW3.5614$JD1.374603@typhoon.nyroc.rr.com>

If comma delimited files are what your after then I have some 'C' code for
you.

This came up on the Zope list a week ago and there was a regular expression
solution.
Li Dongfeng  "Re: [Zope] Comma delimited file and the re python object"

"""
Tests for splitQuote.pyd
"""
from splitQuote import *
import time

txt=r'''
"Bond, James", 13, 45
'That\'s "absolutely","possible."', 1999, 8, 30
"""now"",""is"\'"",the,555,"""time"""
'''
#"""

print split.__doc__
buf=txt*2
print 'Buf len:',len(buf)
t1=time.time()
l=split(buf)
print 'Num records:',len(l)
print "Time: %.2f"%(time.time()-t1)
for x in range(4):
    print l[x]
valid=[['Bond, James', '13', '45'],
['That\\\'s "absolutely","possible."', '1999', '8', '30'],
['""now"",""is"\\\'"', 'the', '555', '""time""'],
['Bond, James', '13', '45'],
['That\\\'s "absolutely","possible."', '1999', '8', '30'],
['""now"",""is"\\\'"', 'the', '555', '""time""']]
assert(l==valid)

--
--Darrell

Roger Irwin <irwin at mail.com> wrote in message
news:942231502.1914033961 at news.interbusiness.it...
> Is there any function or module available for parsing ASCII delimited
files,
> before I go and re-invent the wheel writing my own.
>
> BTW, I know it is simple to do, but I was thinking along the lines that a
> 'standard' could easily be substituted for a C version for large files.
>






From pereira at research.att.com  Tue Nov  2 23:27:52 1999
From: pereira at research.att.com (Fernando Pereira)
Date: Wed, 3 Nov 1999 04:27:52 GMT
Subject: A Comparison of Python and Ruby Extension Interfaces (Re: Comparison between Python and "Ruby")
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp> <7vjtrc$hmm$1@newshost.accu.uu.nl> <38204614.2EAC6051@compaq.com>
Message-ID: <021119992327527678%pereira@research.att.com>

In article <38204614.2EAC6051 at compaq.com>, Greg Ewing
<greg.ewing at compaq.com> wrote:
> * Since Ruby has mark-and-sweep garbage collection, there
> is no manipulation of reference counts. There weren't
> actually any examples of this in the Python dbm module,
> though.
What's the effect of this on portability? For instance, how are
pointers to live objects from the C stack handled? As is well-known,
this cannot be done portably.
> 
> * Ruby seems to be able to raise exceptions without
> having to explicitly propagate them back out of routines,
> which eliminates a lot of result testing and returning of
> -1's and None's that goes on in Python extension code.
Same comment. There is no portable way of unwinding the C stack. What
does that mean for Ruby's portability?

These were pretty ugly problems when I devoted quite a bit of effort to
them in the 80s (Prolog<->C integration back then), and they can only
have become worse with new architectures and more highly optimizing C
compilers.

-- F



From robin at alldunn.com  Thu Nov 18 14:35:32 1999
From: robin at alldunn.com (Robin Dunn)
Date: Thu, 18 Nov 1999 11:35:32 -0800
Subject: Freezing wxPython apps
References: <80v14b$b1s$1@nnrp1.deja.com>
Message-ID: <GyYY3.35$xT.4391081@news.randori.com>

yinger <ayinger1 at my-deja.com> wrote in message
news:80v14b$b1s$1 at nnrp1.deja.com...
> Is there a way to 'freeze' a wxPython app so that it can be distributed
> without requiring wxPython to be installed on the client?
>

I havn't tried freeze on a wxPython app yet, but there shouldn't be a
problem so long as you distribute the binary extension modules too.  For
Win32 though, Gordon's tool is definitly the way to go, as Carrie mentioned.


--
Robin Dunn
Software Craftsman
robin at AllDunn.com
http://AllDunn.com/robin/
http://AllDunn.com/wxPython/  Check it out!






From ivanlan at callware.com  Sun Nov  7 21:21:37 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Sun, 07 Nov 1999 19:21:37 -0700
Subject: pythonistas or pythoneers?
References: <07111999.2@sanctum.jae>
Message-ID: <382633B1.50993D93@callware.com>

Hi All--

"Juergen A. Erhard" wrote:
> 
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> I'm confused: are we pythoneers or pythonistas?
> 
> I tend to go for pythoneers... sounds more pythonic.
> 
> Or are these two levels of experience?  If so, what's higher?
> 

Be confused no longer; as we are all guerilla programmers, we are of
course Pythonistas!

<come-the-revolution>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From billtut at microsoft.com  Wed Nov 24 09:59:36 1999
From: billtut at microsoft.com (Bill Tutt)
Date: Wed, 24 Nov 1999 06:59:36 -0800
Subject: COM
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8100DB915AC@RED-MSG-50>

app.Quit()
del app

Will do the trick.

Bill

> -----Original Message-----
> From: Le Guennec [mailto:jy.lg at wanadoo.fr]
> Sent: Thursday, November 18, 1999 5:22 AM
> To: python-list at python.org
> Subject: COM
> 
> 
> I've created a app variable :
> app = win32com.client.dispatch('Excel.Application')
> 
> how can I free this object and close Excel properly?
> 
> Thanks
> 
> 
> 
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 



From aahz at netcom.com  Thu Nov 25 00:43:33 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 25 Nov 1999 05:43:33 GMT
Subject: CRC-module
References: <000801bf36c8$92ee19a0$292d153f@tim>
Message-ID: <81iia5$pj5$1@nntp9.atl.mindspring.net>

In article <000801bf36c8$92ee19a0$292d153f at tim>,
Tim Peters <tim_one at email.msn.com> wrote:
>
>In practice, I like a two-step approach:
>
>1. Compute CRCs.  If they're different, the files are definitely different,
>   and you get out cheap.
>
>2. Compute SHA-1.  If they're different too, the files are also definitely
>   different.  If they're the same, then since the CRC and SHA-1 methods
>   are unrelated, it's reasonable to expect that the odds of a false match
>   are about 1 in 2**32 * 2**160 == 2**192 now.  This is even smaller than
>   the chance Python2 will have curly-brace block delimeters.

In practice, the likelihood of two files having the same MD5 or SHA is
nearly infinitesimal, provided they're the same length.  Check the first
hundred or thousand characters for equality and you're essentially
done.
--
                      --- Aahz (@netcom.com)

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

Question authority: ask me anything!



From janssen at parc.xerox.com  Tue Nov 30 22:31:47 1999
From: janssen at parc.xerox.com (Bill Janssen)
Date: Tue, 30 Nov 1999 19:31:47 PST
Subject: '==' vs. 'is' behavior 
In-Reply-To: Your message of "Mon, 29 Nov 1999 21:13:40 PST."
             <slrn846n92.kpg.nascheme@adler.dynodns.net> 
Message-ID: <99Nov30.193151pst."3757"@watson.parc.xerox.com>

Actually, the presence of "is" and "==" in Python is one of the nicer
features.  "is" is the Lisp "eq", while "==" is the Lisp "equal".
These are really the two comparisons from Lisp that make the most
sense.

Java, on the other hand, only gives you "eql"!  Ugh!  Bletcherous!

Bill



From phd at phd.russ.ru  Thu Nov 25 04:56:36 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Thu, 25 Nov 1999 09:56:36 +0000 (GMT)
Subject: Listing a dictionary
In-Reply-To: <383CFEE5.951376A0@obop.com.pl>
Message-ID: <Pine.LNX.4.21.9911250956200.11528-100000@fep132.fep.ru>

On Thu, 25 Nov 1999, [iso-8859-2] Przemys?aw G. [iso-8859-2] Gawro?ski wrote:
> How can I list all the keys in a dictionary ?

   dict.keys()

Oleg.
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From parkw at better.net  Fri Nov 19 08:02:02 1999
From: parkw at better.net (William Park)
Date: Fri, 19 Nov 1999 08:02:02 -0500
Subject: Python Chinese
In-Reply-To: <19991119121544.5006.qmail@hotmail.com>
References: <19991119121544.5006.qmail@hotmail.com>
Message-ID: <19991119080202.A439@better.net>

On Fri, Nov 19, 1999 at 01:15:44PM +0100, Ulf Engstr?m wrote:
> Is there any way to use Python with Chinese and other asian chars?
> If not, are there any plans to make this compability?
> Best regards
> Ulf Engstr?m
> 
> http://www.b2b-link.com
> ulf.engstrom at b2b-link.com

What do you mean by "compability"?  All Python has to do is handle
strings in 8-bit clean fashion, then you can cut, paste, and search
8-bit characters as you like.  Unfortunately, I can't find out right
now.

William Park



From a.eyre at optichrome.com  Wed Nov 24 04:29:11 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Wed, 24 Nov 1999 09:29:11 -0000
Subject: Embedded Python 1.5.1 crashes under WIN32
In-Reply-To: <3832F6FB.68AAA96E@tu-bs.de>
Message-ID: <002901bf365e$5dd9e510$3acbd9c2@peridot.optichrome.com>

> What I want to do is to call PyRun_SimpleFile from my main program. This
> works fine under Irix 6.5, but the same thing crashes inside
> PyRun_SimpleFile when compiled with VC6.0SP2 under Win32, which is a
> great pity.

Make sure you aren't mixing runtime DLLS.
i.e. only build debug (MSVCRTD) code with the debug lib (python15_d.lib)
and same for release (MSVCRT + python15.lib)

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From simon at george.maths.unsw.edu.au  Tue Nov 30 20:07:49 1999
From: simon at george.maths.unsw.edu.au (Simon Evans)
Date: 1 Dec 1999 01:07:49 GMT
Subject: Plotting a single pixel (possible newbie question)
Message-ID: <821sd5$7h0$1@mirv.unsw.edu.au>

   I'm relatively new to Python, which I use for several
 small numerically-oriented projects that are refreshingly
 different from my workday giant Fortran simulations.

   I've installed and used both Tkinter and DISLIN (a 
 scientific graphing extension), but what I'm really missing
 is the ability to do the simplest graphics task of all:
 lighting up a single pixel. You'd think it wouldn't
 be too much to ask...

   Does anyone know of a graphics extension that offers this
 simple facility, or a method of doing it in the ones I have?
 I'd like to do it by some other method than drawing a line
 with a length of one pixel!

=================================================================
Simon Evans (simon at maths.unsw.edu.au)
Physical Oceanography Group
School of Mathematics, University of New South Wales, Australia.
=================================================================




From greg.ewing at compaq.com  Thu Nov  4 10:38:08 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Fri, 05 Nov 1999 04:38:08 +1300
Subject: Some more Ruby observations (Re: Comparison between Python and "Ruby")
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp> <7vjtrc$hmm$1@newshost.accu.uu.nl>
Message-ID: <3821A860.5D32AFCC@compaq.com>

Here are some random observations from studying the Ruby
source code:

* It seems to have a notion of "tainted" strings, and can
  be run in various security modes where certain operations
  using tainted strings are disallowed.

* It doesn't use bytecode. The evaluator seems to operate
  directly on a parse tree. I don't know what implications
  this has for execution speed.

* Here is an interesting little diagram I came across. Note
  that Ruby appears to have metaclasses of some flavour or
  other.

    /*
     * Ruby's Class Hierarchy Chart
     *
     *                           +------------------+
     *                           |                  |
     *             Object---->(Object)              |
     *              ^  ^        ^  ^                |
     *              |  |        |  |                |
     *              |  |  +-----+  +---------+      |
     *              |  |  |                  |      |
     *              |  +-----------+         |      |
     *              |     |        |         |      |
     *       +------+     |     Module--->(Module)  |
     *       |            |        ^         ^      |
     *  OtherClass-->(OtherClass)  |         |      |
     *                             |         |      |
     *                           Class---->(Class)  |
     *                             ^                |
     *                             |                |
     *                             +----------------+
     *
     *   + All metaclasses are instances of the class `Class'.
     */

Greg



From mal at lemburg.com  Thu Nov  4 03:59:09 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Thu, 04 Nov 1999 09:59:09 +0100
Subject: Module for generic platform information: platform.py
References: <3815782E.FC516671@lemburg.com>
Message-ID: <38214ADD.1C7AB1CD@lemburg.com>

There is a new release of the platform.py module available at:

        http://starship.skyport.net/~lemburg/platform.py

To see if it works on your platform, simply run the script.

Many thanks go out to all those who have helped so far. The list
of supported platforms is constantly growing :-)

Note that this module is a fast moving target. I plan to release
version 1.0 as the final version.

Still needed:
    ? support for WinCE
    ? support for MS-DOS (PythonDX ?)
    ? support for Amiga, VMS and other platforms running Python
    ? support for additional Linux distributions
    ? debugging of the WinXX code and the JPython code for Java

The module now includes support for Win95,98,2000,NT (still needs
some debugging though), MacOS, MacOS X Server, Linux (RedHat,
Suse, Debian, Slackware, Mandrake), FreeBSD, Solaris,
HP UX, IRIX, OSF (DEC Digital Unix), Java, plus probably some
others that I've forgotten in the list...

Enjoy,
--
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                    57 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/




From donn at u.washington.edu  Tue Nov  2 12:48:56 1999
From: donn at u.washington.edu (Donn Cave)
Date: 2 Nov 1999 17:48:56 GMT
Subject: __str__ vs. __repr__
References: <19991102073422.B134170008@vislab.epa.gov> <1270634443-26149069@hypernet.com> <000701bf24ef$e2d29b20$0f2d153f@tim>
Message-ID: <7vn868$p18$1@nntp6.u.washington.edu>

Quoth Randall Hopper <aa8vb at yahoo.com>:
...
| Thanks.  That's the source of my confusion.  I wanted to follow the
| language convention.  But Python's internals currently don't follow it, so
| the point is somewhat moot.
|
| For consistency, would it make sense to change this for Python 1.5.3 (that
| is, have sequence and dict types pass 'str-vs-repr'ness down)?

I'd go back to an earlier followup on this and try to refine the
critical point, why __str__?  The man said __str__ is intended to
be ``friendly''.

I think it might help to be a little more specific about whose friend
this string is.  For me, anyway, __str__ is for an object whose data
model supports a string value.  For example, if we toss a TypeError
exception value into a context where a string is expected, it might
act like a string 'illegal argument type for built-in operation'.
This isn't just an abridged representation.  In a sense it doesn't
represent the object at all directly, but rather allows the object
to substitute a data value that the programmer expects will be useful,
based on some insight into the meaning of the object and its data
in some context.  In this context, that string value is friendly, but
in any other context, it's deceptive and inappropriate.  In the
distribution library, you'll find lots of __repr__ methods and
relatively few __str__ methods, I guess because it's relatively
unusual to find a string value that's really friendly in a broad
enough context for a library class.

Now there is no such accounting for the context in which sequences
and dicts will be used, they're used everywhere.  The string value
of these objects should directly represent them as the kind of objects
they are, and I think it follows that in this context their contents
should be rendered in the same literal way - to substitute __str__
values instead would only compromise that representation.

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu



From kuncej at mail.conservation.state.mo.us  Mon Nov 15 12:43:15 1999
From: kuncej at mail.conservation.state.mo.us (Jeffrey Kunce)
Date: Mon, 15 Nov 1999 11:43:15 -0600
Subject: DBF interface
Message-ID: <s82ff484.078@mail.conservation.state.mo.us>

>Is there a DBF file interface module ?

Yet another python dbf interface.
    dbfpy.tgz
at
    http://starship.python.net/crew/jjkunce/    


  --Jeff





From gawron at obop.com.pl  Mon Nov 29 08:00:49 1999
From: gawron at obop.com.pl (=?iso-8859-1?Q?Przemys=B3aw?= G. =?iso-8859-1?Q?Gawro=F1ski?=)
Date: Mon, 29 Nov 1999 14:00:49 +0100
Subject: mktime() ???
Message-ID: <38427901.10B1965D@obop.com.pl>

Hello

In library reference it's writen, that mktime from time module raises an
exeption when it cann't construct a correct time from a tuple parameter.

I tryied the following:

mktime((1999, 23, 29, 0, 0, 0, 0, 0, 0 ))

(the month number is incorect),

and

mktime((1999, 23, 49, 0, 0, 0, 0, 0, 0 ))

(the month, and day aren't correct).

Well, neither one of them has raised an exeption.

Is it a bug or am'I not understanding somthing in the docs ?

Thanks for help Przemek

--
Przemyslaw G. Gawronski UIN:8358522
mailto:gawronskip at usa.net
mailto:gawron at obop.com.pl





From usenet-mm-yyyy at julesallen.com  Sun Nov 14 15:56:09 1999
From: usenet-mm-yyyy at julesallen.com (Jules Allen)
Date: Sun, 14 Nov 1999 20:56:09 GMT
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us>
Message-ID: <JlFX3.3659$s87.67930@dfiatx1-snr1.gtei.net>

In article <3d1z9sq2uc.fsf at amarok.cnri.reston.va.us>, Andrew M. Kuchling wrote:
| The issue of indexing modules is an old one, and I'm still astounded
| that we've never converged on a solution, though everyone always
| complains about this.  

Thanks for the History lesson! This helps put it all in perspective
but why *something* hasn't been built before now is kind of weird.
The best part of your story for me is confirmation that open source
projects don't do well without strong leadership. No ESR or an
ESR-like figure to crack the whip, no Trove.

| http://www.vex.net/~x/parnassus/ is *very* nice, and it might be the
| only solution required.  Apparently a good deal of work has been put
| into seeding the initial database, and that's important, because it
| makes the index immediately useful.  Yet I hadn't heard about it until
| now; is the site still beta?  Did I miss a c.l.py.announce article
| about it?  

The author was still tweaking it when he contacted me after my
initial post. I urged him to post the URL for further inspection.
But that's not necessary now as it's been outed!

Monday morning will be a good test of how robust it really is!

By the looks of the vex.net server (Apache/1.3.6 (Unix) PHP/3.0.7),
this could well be a straight cgi app so go easy on it.

| We need to just pick a solution and stick with it; Parnassus is as
| good as any, and has the incalculable advatange of being implemented.

See the lazy bit above. I think it's fabulous and it's bookmarked
and already soothing my info itch. 

Cheers,
Jules

-- 
Want to reply via email? Put the current month and year in 
my email address.



From ivanlan at callware.com  Tue Nov  2 14:14:02 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Tue, 02 Nov 1999 12:14:02 -0700
Subject: Barf
Message-ID: <381F37FA.3899AAEF@callware.com>

Hi All--
I just noticed something rather interesting.  I ran this script:

----------cut here-----------------
#!/usr/local/bin/python

from Tkinter import *
import tkMessageBox
import sys

def die():
    sys.exit(0)

def callee():
    print "I was called; few are chosen"

def about():
    tkMessageBox.showinfo("tkmenu","This is tkmenu.py Version 0")
    
root = Tk()
bar = Menu(root)

filem = Menu(bar)
filem.add_command(label="Open...", command=callee)
filem.add_command(label="New...", command=callee)
filem.add_command(label="Save", command=callee)
filem.add_command(label="Save as...", command=callee)
filem.add_separator()
filem.add_command(label="Exit", command=die)

helpm = Menu(bar)
helpm.add_command(label="Index...", command=callee)
helpm.add_separator()
helpm.add_command(label="About", command=about)

bar.add_cascade(label="File", menu=filem)
bar.add_cascade(label="Help", menu=helpm)

root.config(menu=bar)
root.mainloop()
----------cut here-----------------

It works fine, but if I move the mouse pointer out of the info box
(select Help->About) and back to the main window, I get a segmentation
violation.

This is running on Windows NT 4.0, SP 5; Dell 400MHZ PII Cpu, 128MB
ram.  It does *not* happen on Windows 98, running the identical version
of Python 1.5.2 (latest download, but not CVS).

Also, it makes no difference whether I run it 'python tkmenu.py' or just
'tkmenu.py' (using MKS toolkit).

Is this a known bug?

<there-ain't-no-bugs-on-me>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From tony at lsl.co.uk  Wed Nov 17 05:12:06 1999
From: tony at lsl.co.uk (Tony J. Ibbs (Tibs))
Date: Wed, 17 Nov 1999 10:12:06 -0000
Subject: ANN: Metalanguage (v0.3) for mxTextTools
Message-ID: <0e7401bf30e4$344f1d20$f0c809c0@lslp7o.lsl.co.uk>

This is to announce v0.3 (tiepin) of my metalanguage translator for
mxTextTools.

Marc-Andr? Lemburg provides mxTextTools, a package of fast text manipulation
tools for use with Python. The tagging engine within that package takes as
input a set of tuples, the Tag Table, which tells it what to do. Although
simple in format, these tuples aren't particularly readable, and indeed
Marc-Andr? himself says:

"I admit, these tables don't look very elegant. In fact I would much rather
write them in some meta language that gets compiled into these tables
instead of handcoding them. But I've never had time to do much research into
this, so unless someone volunteers, it will stay like this for some time."

This version of the metalanguage translator implements enough notation to
allow its own tag tables to be written in the metalanguage. Although it
doesn't yet support various things (notably, strings with escaped quotation
marks, function calls, and embedded tables with T/F conditions) it already
provides enough power for most of my own needs, and should hopefully be
useful to other people.

Significant advantages of using the metalanguage are:
* it avoids the problem of counting (and recounting) offsets between tuples
* it avoids the pitfall of forgetting a comma between two tuples (which on
   my Windows machine at home is enough to crash Python)
* it enforces indentation for meaning in the same way that Python itself
   does
* it provides an "if" construct
* obviously, I believe it to be more readable than the bare tuple notation

Licensing: I like the Python Powered license at
     http://starship.python.net/~just/pythonpowered/
so please apply that with suitable amendments. Although Marc-Andr? is aware
of this work, it is in no way his fault (apart from the quote above and some
helpful comments, for which thanks), so don't blame him...

Note that mxTextTools v1.1.1 or later is required to use the translator.

The metalanguage description and translator is available at
    http://www.tibsnjoan.demon.co.uk/mxtext/Metalang.html

mxTextTools itself is available at
    http://starship.python.net/crew/lemburg/mxTextTools.html

<P><A
HREF="http://www.tibsnjoan.demon.co.uk/mxtext/Metalang.html">Metalanguage
v0.3 for mxTextTools</A> - A little language (and its translator) to make it
easier to write mxTextTools tag tables. (16-Nov-99)

--
Tony J Ibbs (Tibs)      tony at lsl.co.uk  http://www.tibsnjoan.demon.co.uk/
Feet first with 5 wheels...
My views! Mine! Mine! (Unless Laser-Scan ask nicely to borrow them.)




From jam at quark.emich.edu  Sun Nov 21 03:42:45 1999
From: jam at quark.emich.edu (Jeff)
Date: Sun, 21 Nov 1999 03:42:45 -0500
Subject: Order of keyword arguments
In-Reply-To: <38379233.ADE65CDB@corrada.com>
References: <38379233.ADE65CDB@corrada.com>
Message-ID: <19991121034245.A21255@quark.emich.edu>

On Sun, Nov 21, 1999 at 01:33:23AM -0500, Andres Corrada wrote:
> Hi,
> 	Can one access the order in which keyword arguments were passed to a
> function? I want to call a function with the keyword syntax: f(
> foo='theFooString', bar='theBarString' ) and use the fact that foo was
> listed first so I can do something like:
> 
> def f( **kw ):
> 
> 	Process keyword arguments remembering the order in which the were used
> to call the function.
> 	Print first keyword
> 	Print second keyword
> 	.
> 	.
> 	.
> 

keyword arguments are passed as a dictionary, and thus the 'order' of the
arguments will not be preserved. what is the advantage you see to being able
to obtain that information? wouldn't a redesign of the interface keep things
simpler and increase the likelihood of success?

hope that helps.

regards,
J
-- 
|| visit gfd <http://quark.emich.edu/>
|| psa member #293 <http://www.python.org/> 
|| New Image Systems & Services, Inc. <http://www.newimage.com/>



From Gaetan_Corneau at baan.com  Tue Nov 30 14:06:19 1999
From: Gaetan_Corneau at baan.com (Gaetan Corneau)
Date: Tue, 30 Nov 1999 14:06:19 -0500
Subject: WinNT and python as text filter...
Message-ID: <816010E2456BD111A48700805FBBE2EEFDF877@ex-quebec-u1.baan.com>

> >If so, there is a bug in the NT command prompt - you need to 
> specify the
> >full path to Python.exe on the command line, and it will work...

This is exactly the problem I have reported a few months ago.
I have access to many tests machines, and two of them had this problem from
time to time.
After a reboot, it disappeared, and I haven't seen it again since :)
______________________________________________________
   Gaetan Corneau
   Software Developer (Quality Assurance Team)
   BaaN  Supply Chain Solutions  
   http://www.baan.com
   E-mail: Gaetan_Corneau at baan.com                  
   Tel: (418) 266-8252          
______________________________________________________
"Profanity is the one language all programmers know best"




From karanvasudeva at usa.net  Sun Nov  7 00:31:51 1999
From: karanvasudeva at usa.net (Karan Vasudeva)
Date: Sun, 07 Nov 1999 05:31:51 GMT
Subject: exec / execfile namespace problem
References: <000b8d9b.b9d3bdca@usw-ex0101-003.remarq.com> <1270182562-5943662@hypernet.com>
Message-ID: <8032s8$jk4$1@nnrp1.deja.com>

Thanks a lot for the advice. I guess since I do know what the code
is doing, I really should structure it that way.

Thanks, y'all.
Karan.

In article <1270182562-5943662 at hypernet.com>,
  gmcm at hypernet.com wrote:
> Karan Vasudeva has problems with exec:
>
> > .... my
> > problem is, when i call my script from another script (by doing
> > an exec, thats what i need to do, you see), the interpreter stops
> > at httplib.py as it can't find that socket called self.sock .
> > This appears to be a namespace issue. I just want to know how to
> > do this right.
>
> exec is pretty much only used when you don't know what
> you'll be running at the time you wrote it. If you know what
> you're running at the time you write it, there's no reason to use
> exec. You probably just want:
>
>   import cool
>
> and maybe:
>   cool.main(args)
>
> Many python scripts have the form:
> -----------------
> def main():
>   # whatever
>
> def otherthing():
>   # stuff used by main, or usable by others
>
> if __name__ == '__main__':
>   main() # if invoked as a script, run main()
> ---------------------------------
>
> Then they can run as the top level script, or they can be
> imported, and any functions, classes etc. used by the
> importer.
>
> - Gordon
>
>


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



From alex at somewhere.around.here  Wed Nov 24 23:51:36 1999
From: alex at somewhere.around.here (alex)
Date: 25 Nov 1999 15:51:36 +1100
Subject: preprocessor for Python?
References: <aBY_3.23149$V4.277574@news1.frmt1.sfba.home.com>
Message-ID: <bvwpuwzcgx3.fsf@room.mathdom.anu.edu.au>

> I'm looking for some macro facilities so I can include debug
> statements that are removed when I need to test performance.

There was a post a while back where it was stated that the __debug__
variable is evaluated at compile time, and blocks beginning with

if 0: 

are not compiled.  So if you set __debug__=0, then any block beginning
with 

if __debug__:

will not be compiled into the program.  You can probably get the
straight dope by searching for __debug__ on deja.

HTH.
Alex.



From steve at acer.spvi.com  Tue Nov  2 17:20:28 1999
From: steve at acer.spvi.com (Steve Spicklemire)
Date: Tue, 2 Nov 1999 17:20:28 -0500 (EST)
Subject: Python director XTra....  cute 'robot' toy for mac
In-Reply-To: <199910291511.KAA26339@acer.spvi.com> (message from Steve
	Spicklemire on Fri, 29 Oct 1999 10:11:55 -0500 (EST))
References: <199910291511.KAA26339@acer.spvi.com>
Message-ID: <199911022220.RAA21086@acer.spvi.com>

There's a 'start' of a 'robot' toy for the mac at:

http://opensource.spvi.com

It's a macromedia projector that imports a module (called robots.py)
in the movie's folder and uses it to calculate the trajectory of a
sprite in the movie. This is really crude, but shows a 'proof of
concept' that could be really cool....  The windows version is also
(almost) ready. You have to have MacPython installed for this to work.

-steve



From gerrit.holl at pobox.com  Mon Nov  1 14:31:20 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Mon, 1 Nov 1999 20:31:20 +0100
Subject: C-Python tutorial?
Message-ID: <19991101203120.A29797@optiplex.palga.uucp>

Hello,

I have learned python as my first langauge some time ago. Is there a C
tutorial for python programmers, that covers C programming, espaccialy for
extending it with python?

I think not. Can anyone write one? I would be very grateful!

regards,
Gerrit.
-- 
linuxgames.nl.linux.org	 All about games on Linux in Dutch (under construction).
www.nl.linux.org	       The Dutch resource for Dutch Linux documentation.
www.nl.linux.org/discoverb		    Learn foreign words and definitions.
www.nl.linux.org/~gerrit/asperger	 I'm an Asperger, go here for more info.
www.nl.linux.org/~gerrit			    About me and pictures of me.



From pjensen at columbus.rr.com  Thu Nov  4 01:32:50 1999
From: pjensen at columbus.rr.com (Philip H. Jensen)
Date: Thu, 4 Nov 1999 01:32:50 -0500 (EST)
Subject: Python and Literate Programming
Message-ID: <Pine.LNX.3.96.991104011425.943A-100000@dhcp9545112>

Greetings.  I'm a recent convert to Python, and wish I had looked into
it years ago.  As I said in a posting on my local Linux UG's mailing
list, "I like everything about Python, whereas there are many things
about Perl that I don't like."

Incidentally, I'm also a booster for Modula-3, and of course Python
has well-known and intentional similarities to M3.  But one of my other
enthusiasms is Literate Programming, and I'm wondering about the
possibilities for it in Python.

Documentation texts can be easily accomodated, as can pretty-printing.
But for me, the core of LP, which I use (e.g. in CWEB) even if I never
run the program through CWEAVE, is >> factoring <<.

Here's how it might look in a Python-friendly form.

	if not dictionary.has_key(word):
	    <Inform the user that |word| is not present in
		the dictionary, and take action as directed>

then elsewhere in the program:

	<Inform the user that |word| is not present...>:
	    code...

***
Now, personally I would like to see this in the language itself,
but that seems unlikely.  What would be essential, though would be

  1) a hook whereby "import foo" would see if there was a "foo.pyw"
     more recent than "foo.py", and if so, would invoke a pytangle
     program, and
  2) a means for file and line information to be passed along, so
     that error messages would refer to the .pyw source.

Any interest?  If I myself were to work on patches along these
lines, could they become candidates for inclusion in the mainstream
source tree?

    (I have some other Python development ideas,
     but will save them for later postings.)
  Phil Jensen
  pjensen at columbus.rr.com 




From x at x.x  Mon Nov 29 17:35:36 1999
From: x at x.x (Eric Jacobs)
Date: Mon, 29 Nov 1999 17:35:36 -0500
Subject: performance of tuple-less Python?
References: "Greg Wilson"'s message of "Mon, 29 Nov 1999 11:29:30 -0500" <1268223702-30481186@hypernet.com>
Message-ID: <3842FFB8.C2146DCE@x.x>

Gordon McMillan wrote:
>
> I'm guessing that not being able to find something you stuck
> in the dictionary qualifies as a performance penalty ;-).
> 
> To spell it out:
>  1) hash based on id(list_object):
>      d = {[1] = 'x'}
>      d[[1]] -> KeyError

Actually, this behavior can be useful at times. I was working on a graphviz
grapher to plot Python's internal ASTs. If I converted them to lists, Python
would complain becauses lists are not hashable. But when I used tuples, I saw
that Python was cleverly performing common subexpression analysis. So that in

  f(x + 1, x + 1, x + 1)

there would be only one subtree representing "x + 1", pointed to three times.
I had to explicitly state the id() operation to get it to the way (I thought)
it should look.

>  2) hash based on value:
>      lst = [1]
>      d[lst] = 'x'
>      lst.append(2)

The penalty would be that the dictionary would have to be reindexed at this
point, because the hash and compare behavior of lst would have changed.

>      d[lst] -> KeyError
> 
> - Gordon



From guido at cnri.reston.va.us  Wed Nov  3 17:24:55 1999
From: guido at cnri.reston.va.us (Guido van Rossum)
Date: Wed, 03 Nov 1999 17:24:55 -0500
Subject: Python Conference Hotel Change
Message-ID: <199911032224.RAA10668@eric.cnri.reston.va.us>

The Foretec conference staff has this message about the upcoming
Python conference:

Due to construction delays, the Hilton Alexandria Old Town will not be
completed in time for the Eighth International Python Conference.
However, we are pleased to announce that we have secured an equally
desirable facility for the conference.  The new facility is the Key
Bridge Marriott in Arlington, Virginia.  The Key Bridge Marriott is
conveniently located on the Virginia side of the Key Bridge in the
Rosslyn business district with easy access to restaurants, shops, and
the Rosslyn metro stop.  It overlooks the Potomac River, and is a 15
minute walk across the bridge to historic Georgetown in Washington, DC.

(I should add that Georgetown is, if possible, an even better location
than old town Alexandria for the typical Python conference evening
activities. --Guido :-)

--Guido van Rossum (home page: http://www.python.org/~guido/)



From billtut at microsoft.com  Wed Nov 10 07:38:00 1999
From: billtut at microsoft.com (Bill Tutt)
Date: Wed, 10 Nov 1999 04:38:00 -0800
Subject: Using GNU C complier as the "default" windows complier
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8100DB9151B@RED-MSG-50>

cygwin isn't thread safe at this moment in time.
If anything, Python should be compiled using mingw32 instead of cygwin.
Esp, in order to get the Win32 extensions to be useful.

However.... I have no idea if the Win32 extensions actually can be compiled
successfully using mingw32 as a substantial portion of them are C++, and
even if they could be compiled, there would be binary incompatability
problems due to using C++. (C++ mangling is different for all C++
compilers...) So given those two things, and neither Mark, I, or anybody
else atm seems to care enough to bother trying to see if it'll work. The
binary compatability problem is the biggest problem though. If a MSVC
compiled Python install can't make use of a mingw32 Win32 extension DLL (or
vice versa) then even bothering with trying becomes counter productive.
(unless your willing to invest time in submitting changes to the source code
to remove the C++ mangling issues) i.e. Turn all exported C++ objects into
COM objects or something so you can just use COM vtbls to communicate.
Needless to say, thats a lot of work that I'd rather not do. (I expect Mark
won't be to gung ho to do that work either... :) )

Bill

> -----Original Message-----
> From: Werner Fouche [mailto:wfouche at mindspring.com]
> Sent: Wednesday, November 10, 1999 3:59 AM
> To: python-list at python.org
> Subject: Re: Using GNU C complier as the "default" windows complier
> 
> 
> 
> "Mike Callahan" <mcalla at home.com> wrote in message
> news:ma4W3.7794$Ur4.223420 at news.rdc1.tn.home.com...
> > This a question from a newbie but I was thinking...why 
> doesn't the default
> > Python code for windows use the GNU C complier for windows 
> instead of the
> MS
> > complier. Would it be possible to distribute the GNU 
> package along with
> the
> > Python package and thus make it easier to compile, extend, and embed
> Python,
> > use freeze, etc? I know it would take some work, but it 
> seems it would
> make
> > windows Python a lot more like unix Python. Just a thought...
> >
> >
> 
> I had the same idea on the weekend. So I installed Cygwin
> (http://www.cygnus.com/cygwin) on my NT system and after running
> "./configure" got Python to compile and run. However, I could 
> not get Python
> threads to work.
> 
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 



From thamelry at vub.ac.be  Wed Nov 24 08:32:21 1999
From: thamelry at vub.ac.be (Thomas Hamelryck)
Date: 24 Nov 1999 13:32:21 GMT
Subject: Python complaints
References: <XFMail.991124113221.mikael@isy.liu.se> <81gj3i$fe3$1@mach.vub.ac.be> <005201bf367e$4901fef0$f29b12c2@secret.pythonware.com>
Message-ID: <81gpd5$332$1@mach.vub.ac.be>

Fredrik Lundh (fredrik at pythonware.com) wrote:
: Thomas Hamelryck <thamelry at vub.ac.be> wrote:
: > I don't want to start the discussion about indentation again; it's no use
: > since indenting blocks is here to stay. The point is that if you want to 
: > write code that looks ugly or unreadable you will succeed in doing that,
: > indentation or not. In many cases, using a variable indentation (depending
: > on what construct you are using or the length of a line) leads to code
: > that is visuably much more appealing.

: you make that sound like a scientific fact.  can
: you point us to some research, or is it just yet
: another random opinion?

Eh...this is my opinion, yes. Do you call everything you don't agree 
with a "random opinion"? Of course it is not a "scientific fact": 
readability of code is in the eye of the beholder to a great extent. 
When it comes to private attributes in python we are all
"consenting adults", but when it comes to code readability the indentation
rules need to be enforced. I think there's a contradiction here. But again,
I am very fond of python despite the indentation rules. 

: (if it is, we simply disagree.  and for the record,
: my opinion is based on many years of experience
: from using Python with casual programmers, and
: from hacking in huge Python programs written by
: other programmers...  what's your opinion based
: on?)

On using python myself (for about a year now) and looking at other people's
 code.  Again I'd like to mention that Python's advantages far outweigh the 
strict indentation rules.

Cheers,

--
Cheers,

------------------------------------------------
Thomas Hamelryck//Free University Brussels (VUB)
Intitute for Molecular Biology// ULTR Department
Paardestraat 65//1640 Sint-Gensius-Rode//Belgium
------------------------------------------------



From nikolai at micon.no  Tue Nov  9 18:19:43 1999
From: nikolai at micon.no (Nikolai Kirsebom)
Date: Wed, 10 Nov 1999 00:19:43 +0100
Subject: MS Word COM error w/ Borders property
References: <38289ED8.9BA2645A@alliedsignal.com>
Message-ID: <pW1W3.5799$1C4.55807@news1.online.no>

Try the following:

for i in range(word.ActiveDocument.Tables(1).Borders.Count):
   print word.ActiveDocument.Tables(1).Borders(i+1).LineStyle

and you'll have all the line styles of the first table in your active
document.

Nikolai






From fdrake at acm.org  Fri Nov  5 11:08:19 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Fri, 5 Nov 1999 11:08:19 -0500 (EST)
Subject: __str__ vs. __repr__
In-Reply-To: <l03102801b44870c20480@[193.78.237.174]>
References: <199911041325.IAA14001@eric.cnri.reston.va.us>
	<l03102801b44870c20480@[193.78.237.174]>
Message-ID: <14371.243.373402.279764@weyr.cnri.reston.va.us>

Just van Rossum writes:
 > As a step towards a Unicode-savvy Python, would it be an idea to define
 > Unicode versions of repr(), str(), ob.__repr__() and ob.__str__()?

  Perhaps it would be more useful for file objects to offer an
attribute or query method that allowed client code to determine if the 
stream supports unicode?


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From a.eyre at optichrome.com  Mon Nov 15 09:27:18 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Mon, 15 Nov 1999 14:27:18 -0000
Subject: [Python] header (fwd)
In-Reply-To: <1269958983-19391623@hypernet.com>
Message-ID: <002a01bf2f75$85166620$3acbd9c2@peridot.optichrome.com>

> Huh? I filter (in Pegasus mail) on "python-list@" in To: and 
> CC: and get them all.

I filter "From:" for "python-list-admin at python.org"

Does that not work?

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From gawron at obop.com.pl  Thu Nov 25 04:18:29 1999
From: gawron at obop.com.pl (=?iso-8859-2?Q?Przemys=B3aw?= G. =?iso-8859-2?Q?Gawro=F1ski?=)
Date: Thu, 25 Nov 1999 10:18:29 +0100
Subject: Listing a dictionary
Message-ID: <383CFEE5.951376A0@obop.com.pl>

How can I list all the keys in a dictionary ?

Thank you

Przemek

--
Przemyslaw G. Gawronski UIN:8358522
mailto:gawronskip at usa.net
mailto:gawron at obop.com.pl





From pinard at iro.umontreal.ca  Mon Nov  1 19:09:33 1999
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 01 Nov 1999 19:09:33 -0500
Subject: Produce .pyc without execution
In-Reply-To: Michael Hudson's message of "01 Nov 1999 21:29:34 +0000"
References: <002001bf244e$8e187600$3acbd9c2@peridot.optichrome.com> <oq904ii2kj.fsf@titan.progiciels-bpi.ca> <m37lk1dhoh.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <oqogddhhz6.fsf@titan.progiciels-bpi.ca>

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

> I think installing .pyc's but not .py's is mischevious, if only
> because meaningful tracebacks are nice.

We have to be careful in our search for virtue.  Richard Stallman always
wanted packages to install un-stripped binaries, for reasons like above.
The disk space it takes, as well as the principle, is not very acceptable
for a lot of users, and Linux packagers defeat it quite systematically,
so far that I know.  It surely required a lot of doing and many fights (I
do know some of them :-) before the thing could be done with more ease than
difficulty within GNU, and I've been told that some of the last problems are
going to be solved in the next release of Autoconf, if not already.  I do
not follow this file closely anymore, now, as helping GNU is too exhausting.

My opinion is that quality packages should not traceback anyway, and
rather produce their own user-oriented diagnostics.  The installation
process should be oriented more towards users than towards debugging,
for anything else than alpha software.

Even for Emacs, Linux packagers set things so LISP sources are optional.
Oh, no doubt that having sources installed is a good things for many users,
but in the long run, we are going to have more and more users who are not
especially programmers, or not especially interested in the inner details
of everything they use, so we should silence our natural proselytism :-).

> > What would be ideal is a way for `compile_dir' to produce `.pyc' in
> > the build hierarchy directly from the source hierarchy at `make all'
> > time, and that `make install' just move these ready `.pyc' files into
> > the installation directories.

> It shouldn't be all that hard to adapt Lib/py_compile.py &
> Lib/compileall.py as appropriate to effect this.

I hope that people responsible for these packages are listening! :-)

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





From mwh21 at cam.ac.uk  Tue Nov 23 18:06:31 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 23 Nov 1999 23:06:31 +0000
Subject: whrandom seed question
References: <81eu22$411al$1@umbc7.umbc.edu>
Message-ID: <m366ysu7t4.fsf@atrus.jesus.cam.ac.uk>

kyriacou at umbc.edu (Stelios Kyriacou) writes:

> Hi, 
> 
> I tried today the random number generator in python 
> and have the following question: 
> 
> Shouldn't the following 3 lines produce the same number? 
> 
> >>> whrandom.seed(0,0,0); print whrandom.random() 
> 0.39613674715 
> >>> whrandom.seed(0,0,0); print whrandom.random() 
> 0.170188890936 
> >>> whrandom.seed(0,0,0); print whrandom.random() 
> 0.831213988577 
> >>> 

Use the source luke!

>From /usr/lib/python1.5/whrandom.py:

        def seed(self, x = 0, y = 0, z = 0):
                if not type(x) == type(y) == type(z) == type(0):
                        raise TypeError, 'seeds must be integers'
                if not (0 <= x < 256 and 0 <= y < 256 and 0 <= z < 256):
                        raise ValueError, 'seeds must be in range(0, 256)'
                if 0 == x == y == z:
                        # Initialize from current time
                        import time
                        t = long(time.time() * 256)
                        t = int((t&0xffffff) ^ (t>>24))
                        t, x = divmod(t, 256)
                        t, y = divmod(t, 256)
                        t, z = divmod(t, 256)
                # Zero is a poor seed, so substitute 1
                self._seed = (x or 1, y or 1, z or 1)

So

>>> whrandom.seed(1,1,1); print whrandom.random()
0.0169309061997
>>> whrandom.seed(1,1,1); print whrandom.random()
0.0169309061997
>>> whrandom.seed(1,1,1); print whrandom.random()
0.0169309061997

as expected.
 
> Am I missing something??? 

New one on me that! You'd think None would be a better default.

Regards,
Michael




From reic0024 at ub.d.umn.edu  Tue Nov 16 10:16:44 1999
From: reic0024 at ub.d.umn.edu (Aaron J Reichow)
Date: Tue, 16 Nov 1999 09:16:44 -0600
Subject: Directory chooser in Tk
In-Reply-To: <14385.23296.680071.39959@gargle.gargle.HOWL>
References: <38304E78.F92C8596@callware.com> <19991116071917.A711753@vislab.epa.gov> <14385.23296.680071.39959@gargle.gargle.HOWL>
Message-ID: <Pine.SO4.4.02.9911160912540.22506-100000@ub.d.umn.edu>


>  >  |I need a dialog similar to the askopenfilename() dialog, but for
>  >  |directories.  Is there such a beast?
>  > 

I needed the same thing.  I just wound up using askopenfilename(),
choosing any file in the directory I wanted, and then using os.path.split()
to dump the name of the file and retain the path in another variable.  Not
elegant, but it works.

Aaron




From akuchlin at mems-exchange.org  Tue Nov 30 15:05:09 1999
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: 30 Nov 1999 15:05:09 -0500
Subject: '==' vs. 'is' behavior
References: <382C2C6A.64A8F2C2@theriver.com> <14380.12036.197083.409565@weyr.cnri.reston.va.us> <8E7F86267duncanrcpcouk@news.rmplc.co.uk> <oqemd9dk2g.fsf@titan.progiciels-bpi.ca> <015301bf3b0a$685a3fc0$f29b12c2@secret.pythonware.com> <oqg0xnapew.fsf@titan.progiciels-bpi.ca>
Message-ID: <3dbt8bhhje.fsf@amarok.cnri.reston.va.us>

Fran?ois Pinard <pinard at iro.umontreal.ca> writes:
> I'm a bit minimalist while programming.  Or, to say the same thing another
> way, I rather like Arvo Part's music.  When a program source is just

In the article at
http://www.itknowledge.com/tpj/issues/vol1_1/tpj0101-0002.html
(excerpted in Linux Journal, Larry Wall writes, as part of an extended
metaphor:

	And languages specialize in other ways: BASIC is like pop
	music. Tune into REXX for your easy listening classics. Tcl is
	fuzzy like jazz - you get to improvise a lot, and you're never
	quite sure who is interpreting what. Python is like MTV - it
	rocks, but it gets to be much of a sameness after half an hour
	or so.

(Out-of-context quoters can have fun by inserting ellipses after "it
rocks". :) ) At the time I thought the Python comparison was a bit
odd, and didn't really see the point.  Python is more like a
collection of tiny interacting modules; no one module is very
complicated, but put them together and they intermesh and interweave
in a beautiful way over time. [1] Over time the modules fall together
in slightly different patterns, and, though the difference between
iterations is slight, the resulting shift can completely change the
effect of the whole piece.

Ergo, Python is a Steve Reich composition. [2]

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
Persichetti could do everything and do it well, but you didn't really care.
    -- Steve Reich, on his former composition teacher Vincent Persichetti

[1] I imagine Lisp and Scheme fans think of programs in those 
    languages in an analogous way.

[2] A comparison to Terry Riley's "In C" would be even closer -- it's
    an open score where each musician has a melody, and they play it as
    often as they like, whenever they like, as long as they don't stray
    too far from the path -- but I like Reich better.




From mcfletch at vrtelecom.com  Tue Nov  2 22:34:21 1999
From: mcfletch at vrtelecom.com (Mike Fletcher)
Date: Tue, 2 Nov 1999 22:34:21 -0500
Subject: Addition to documentation: Zen of Python, was( RE: Comparison between Python and "Ruby" )
In-Reply-To: <381F9EB4.63FA033F@python.net>
Message-ID: <NDBBIAPHEKGOBIHCBLIIAENKCAAA.mcfletch@vrtelecom.com>

Okay, that reminds me, when is this going to make it into the official
documentation?  If it were me, I'd put it right onto the first page
(doc/index.html) of the official docs, right under the listing of available
documents ;) .  It does such a great job of getting people into the "frame"
of Python (at least those with some background in computer programming).

So, how about it? And lets start a Python monastery somewhere, I'd look
great in a black-and-green robe ;) ,
Mike

...
Beautiful is better than ugly.
Explicit is better than implicit.
Simple is better than complex.
Complex is better than complicated.
Flat is better than nested.
Sparse is better than dense.
Readability counts.
Special cases aren't special enough to break the rules.
Although practicality beats purity.
Errors should never pass silently.
Unless explicitly silenced.
In the face of ambiguity, refuse the temptation to guess.
There should be one-- and preferably only one --obvious way to do it.
Although that way may not be obvious at first unless you're Dutch.
Now is better than never.
Although never is often better than *right* now.
If the implementation is hard to explain, it's a bad idea.
If the implementation is easy to explain, it may be a good idea.
Namespaces are one honking great idea -- let's do more of those!
-- Tim Peters




From mwh21 at cam.ac.uk  Mon Nov  1 16:24:12 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 01 Nov 1999 21:24:12 +0000
Subject: __str__ vs. __repr__
References: <19991101121709.A134039715@vislab.epa.gov> <19991101202118.B29736@optiplex.palga.uucp>
Message-ID: <m3aeoxdhxf.fsf@atrus.jesus.cam.ac.uk>

Gerrit Holl <gerrit.holl at pobox.com> writes:

> Randall Hopper wrote:
> > 
> > 1) Is there a convention for what __str__ and __repr__ should return for
> >    classes?
> 
> __str__ should return a string, __repr__ may also return, for example, a
> dictionairy.

Ehh? No, I don't think so:

>>> class Class:
...  def __repr__(self):
...   return {}
...
>>> Class()
Traceback (innermost last):
  File "<stdin>", line 1, in ?
TypeError: repr not string

>From the reference manual (which I could paraphrase, but there'd be no
real point):

   __repr__ (self)
          Called by the repr() built-in function and by string
          conversions (reverse quotes) to compute the ``official'' string
          representation of an object. This should normally look like a
          valid Python expression that can be used to recreate an object
          with the same value. By convention, objects which cannot be
          trivially converted to strings which can be used to create a
          similar object produce a string of the form "<...some useful
          description...>".

   __str__ (self)
          Called by the str() built-in function and by the print
          statement to compute the ``informal'' string representation of
          an object. This differs from __repr__() in that it does not
          have to be a valid Python expression: a more convenient or
          concise representation may be used instead.
 
The idea is that 

eval(repr(x)) == x

should be true as much as is possible, whereas 

str(x)

should be fit for human consumption.

> > 2) Or, whatever they return, should they return the same value?

It's worth noting that the *only* type I know of that has a different
`str' to its `repr' is the string type:

>>> print str("hello")
hello
>>> print repr("hello")
'hello'

Regards,
Michael



From gerrit.holl at pobox.com  Sat Nov 20 07:06:49 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Sat, 20 Nov 1999 13:06:49 +0100
Subject: Naming: fp
In-Reply-To: <slrn82unnr.mja.thor@localhost.localdomain>
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <JlFX3.3659$s87.67930@dfiatx1-snr1.gtei.net> <slrn82unnr.mja.thor@localhost.localdomain>
Message-ID: <19991120130649.A18627@optiplex.palga.uucp>

Hello,

>  A file object fp that  points to the mailbox file.

Why is a file object often called 'fp'?

-- 
"We all know Linux is great...it does infinite loops in 5 seconds."
(Linus Torvalds about the superiority of Linux on the Amterdam
Linux Symposium)



From mikael at isy.liu.se  Fri Nov  5 06:49:08 1999
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Fri, 05 Nov 1999 13:49:08 +0200 (MET)
Subject: Apply a function to each list member?
In-Reply-To: <19991105215504.A497@Ridcully.home>
Message-ID: <XFMail.991105134908.mikael@isy.liu.se>

On 05-Nov-99 Malcolm Tredinnick wrote:
 >  Correct. In fact, since we're being precise, let me point out that
 >  (mathematically speaking), all functions are maps (or mappings), but not
 >  all mappings are functions. The point is that a function should have only a
 >  single "output" for any given input (i.e. it should be many-to-one or
 >  one-to-one, not one-to-many). So some mapping that sends, say, the number 2
 >  to the numbers 3 and 4 is not a function (but a mapping that sends both the
 >  numbers 3 and 4 to the number 2 is a function, since for each input there
 >  is only a single output).

Ahh, there we got it. Good thing someone has got it straight.

 >  (Hmm .. I've just realised that a few of my recent postings have had
 >  amazingly little to do with Python. I promise to stick to the topic from
 >  now one. Swear! :-) )

I-think-I-should-get-in-line-ly y'rs

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    05-Nov-99
Time:    13:47:18

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



From dnagata at creo.com  Sat Nov 13 17:14:46 1999
From: dnagata at creo.com (Dale Nagata)
Date: Sat, 13 Nov 1999 14:14:46 -0800
Subject: finding out console user (Win95)
References: <941814380.235290@cdwork.cvt.stuba.sk>
Message-ID: <382DE2D6.29D1@creo.com>

Radovan Garabik wrote:
> 
> I am porting my script from unix to windows, and I need to find
> console user under win95 (under WinNT I can use os.environ['USERNAME'],
> which is not exactly console user, but good enough for me).
> Something like cuserid() or getlogin() as implemented in
> Cygwin would be ideal....
> Is there a simple way to find it?
> 

The following should work on both Win95 and NT:

	import win32api
	print win32api.GetUserName()

This will get the correct account name even if the USERNAME
environment variable is set to something different.


--
Dale Nagata         |  tel : +1 604.451.2700 ext. 2254 (UTC-0800)
Software Developer  |  fax : +1 604.437.9891 
Creo Products Inc.  |  pgr : +1 604.691.8279
Burnaby BC Canada   |  http://www.creo.com/



From tre17 at student.canterbury.ac.nz  Tue Nov 16 21:37:01 1999
From: tre17 at student.canterbury.ac.nz (Tim Evans)
Date: 17 Nov 1999 15:37:01 +1300
Subject: problems with __getattr__ & __setattr__
References: <3F8BA119B48AD11197470008C7A493CB98B67D@HQOBGEX1>
Message-ID: <c75u2mlzvw2.fsf@pc142.cosc.canterbury.ac.nz>

ScherBi at BAM.com writes:

> I am having trouble using __getattr__ and __setattr__ in some classes.  I've
> boiled it down to the code below.
> It dumps core under 1.5.2 on Linux or WinNT.
> 
> I think maybe it's looping, if so, how does one go about doing this? 
> (I'm assuming it's clear enough what I'm trying to do.)
> 
> If I comment out the __gettattr__ operation, it doesn't dump core.  Instead
> I get an AttributeError: eggs raised at the 'self.eggs[key] = value' call in
> __setattr__.  This is what leads me to belive there's something circular
> going on here.
> 
> 
> Thanks,
>  Bill
> 
> --------------------------------------------
> class spam:
> 
>     def __init__(self):
>         self.eggs = {}
>     
>     def __getattr__(self, key):
>         try:
>             return self.eggs[key]
>         except:
>             raise
>         
>     def __setattr__(self, key, value):
>         try:
>             self.eggs[key] = value
>             return 0
>         except:
>             raise
> 
> if __name__ == '__main__':
> 
>     s = spam()
>     print 'instance of spam created.' 
> -----------------------------------------------

I think I see the problem here.  In __init__ you try to set an
attribute (self.eggs = {}), which will cause __setattr__ to be called.

In __setattr__ you get the eggs attribute, which does not exist, so
the attribute access is passed to __getattr__.

In __getattr__ you access the eggs attribute again, which as it does
not exist will again be passed to __getattr__, causing an infinite
loop.

To demonstrate this add a print statement at the start of each method
that prints the name of the method.  The result is:

-----------------------------
% python broken.py
__init__
__setattr__
__getattr__
__getattr__
__getattr__
__getattr__
__getattr__
__getattr__
__getattr__
...
zsh: segmentation fault  python broken.py
-----------------------------

To solve this problem I would recommend not using __getattr__ and
__setattr__ unless you really need it.  If you don't want to use
s.eggs['foo'] then you could define __setitem__ and __getitem__ to
shorten this to s['foo']

--
Tim Evans



From init__self at my-deja.com  Mon Nov  8 11:21:48 1999
From: init__self at my-deja.com (Python Addict)
Date: Mon, 8 Nov 1999 11:21:48 -0500
Subject: IPC8: Call the 800 number!
Message-ID: <3826f6ac.0@news.bestsoftware.com>

There seems to be a problem between the chairs and the keyboards at the
703-524-6400 Key Bridge Marriott hotel -- the reservation agents don't seem
to know about the conference and were unable to speak to anybody that did.
Therefore they wouldn't honor the $129 rate. I even tried rep fishing but
got the same person again (bah!).

However, the 800 number *does* know about the conference and was more than
happy to take my money. You might want to try them first.





From mhammond at skippinet.com.au  Tue Nov  2 17:43:37 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 3 Nov 1999 09:43:37 +1100
Subject: Reading .lnk files
In-Reply-To: <381F5CD1.8714622F@callware.com>
Message-ID: <004301bf2583$b47a7150$0501a8c0@bobcat>

Sure is.  For more details, see the COM help file on PyIShellLink.

Mark.

# DumpShortcut.py
# A demo of how to dump a shortcut using Python.
 
# This code dumps the shortcut specificed as sys.argv[1]
 
from win32com.shell import shell
import pythoncom
 
def DumpIt(fileName):
# Get the shell interface.
    sh = pythoncom.CoCreateInstance(shell.CLSID_ShellLink, None, \
      pythoncom.CLSCTX_INPROC_SERVER, shell.IID_IShellLink)
     
    # Get an IPersist interface
    persist = sh.QueryInterface(pythoncom.IID_IPersistFile)

    persist.Load(fileName)
     
    # Get the data
    print "Path name:", sh.GetPath(shell.SLGP_SHORTPATH)[0]
    print "Args:", sh.GetArguments()
    print "Description:", sh.GetDescription()
    print "Working directory:", sh.GetWorkingDirectory()

if __name__=='__main__':
    import sys
    DumpIt(sys.argv[1])

# demo of usage:
C:\>DumpShortcut.py IDLE.lnk
Path name: L:\src\PYTHON~1.2\PCbuild\pythonw.exe
Args: l:\src\python-cvs\tools\idle\idle.pyw
Description:
Working directory: L:\src\python-cvs\tools\idle




From piers at cs.su.oz.au  Wed Nov 24 23:10:15 1999
From: piers at cs.su.oz.au (Piers Lauder)
Date: Thu, 25 Nov 1999 15:10:15 +1100
Subject: threading and timeouts
Message-ID: <943504970.076.346131773@cs.usyd.edu.au>

Are timeouts possible inside threads?

Eg: escaping from an attempt to use smtplib.SMTP().sendmail() inside a
child thread if the remote host is being extremely slow.

It doesn't seem possible to use the traditional signal(SIG_ALRM,...)
approach because signals only work inside the main thread.

The only option I can think of is to re-write the smtplib module to use
`select', but does anyone know of a simpler solution?







From jsight at mindspring.com  Tue Nov 30 21:14:58 1999
From: jsight at mindspring.com (Jesse D. Sightler)
Date: Tue, 30 Nov 1999 21:14:58 -0500
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <81bmmi$279$1@nnrp1.deja.com> <3841EE77.6B5D6AC9@wjk.mv.com> <FLxz96.IEI@world.std.com> <38435C43.B238EC15@mindspring.com> <FM0KtD.2ps@world.std.com> <m33dtoro6a.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <384484A2.AF4BE448@mindspring.com>

Michael Hudson wrote:
> 
> wware-nospam at world.std.com (Will Ware) writes:
> 
> > Will Ware wrote:
> > > I find myself grumbling about having to type "x = x + 1". The really
> > > clean thing to do, given that integers are objects, would be to define
> > > increment and decrement methods, so you'd type something like "i.incr()".
> >
> > Jesse D. Sightler (jsight at mindspring.com) wrote:
> > > Wouldn't this be impossible, since Integers are also immutable?  So, eg,
> > > i.incr() could only return an incremented version of i, but not actually
> > > increment i?
> >
> > Right, this is what Gordon was pointing out. A "++" method could not
> > magically reach out of its object, find the "i" variable, and bind it
> > to the incremented value.
> 
> Hate to be awkward, but when it comes to things like this, there's
> very little like this that can't be done in Python!
> 
> >>> def incr(var):
>         try:
>                 raise 1/0
>         except ZeroDivisionError:
>                 import sys,dis
>                 f = sys.exc_traceback.tb_frame.f_back
>                 loadins = f.f_code.co_code[f.f_lasti - 3:f.f_lasti]
>                 loadop = ord(loadins[0])
>                 name = dis.opname[loadop]
>                 loadarg = ord(loadins[1]) + 256*ord(loadins[2])
>                 vname = f.f_code.co_names[loadarg]
>                 f.f_locals[vname] = f.f_locals[vname] + 1
> 
> 
> >>> x=1
> >>> incr(x)
> >>> x
> 2

Ack, and I though Perl was the only language that made you do stuff like
that.  :)  Could you please explain this little piece of code?

It sure looks like byte-code dependant, self-modifying crazyness to me. 
:)

---------------
Jesse D. Sightler
http://www.biddin.com/delorean/



From Gaetan_Corneau at baan.com  Thu Nov 25 10:46:27 1999
From: Gaetan_Corneau at baan.com (Gaetan Corneau)
Date: Thu, 25 Nov 1999 10:46:27 -0500
Subject: Python and COM: a bug? (long)
Message-ID: <816010E2456BD111A48700805FBBE2EEFDF831@ex-quebec-u1.baan.com>

Mark,

I got help and traced the C++ code. Here's a partial  description of the COM
object hierarchy.

ITK.AdapterManager (Root class)
       |
       |
   ITKAdapter      (Used to connect to data sources. Ex: OLEDB)
       |
       |
  ITKConnection    (Actual connection to data source)
       |
       |
    ITKClass       (Meta-data)
       |
       |
  ITKDataObject    (Actual data)


Here is a small example of usage. Suppose tou have an Oracle database you
want to access using OLEDB. In that database, you have an employee table,
with the following fields: Name, address, ID, etc. To get a list of all
employees names, I write:

fromwin32com.client.dynamic import Dispatch

am = Dispatch("ITK.AdapterManager")
ad = am.GetAdapter("OLEDB")
Conn = ad.NewConnection()
Conn.Open("/dsn:HPD380 /user:gc /pwd:qwerty")

for Employee in Conn.Classes("Employee").GetObjects():
    print Employee.Name



What do we use all that for? In an integration server that manipulates
abstract data, reading from a source, writing to a destination. The server
lets you define data magppings graphically (for simple one to one field
mapping). All that stuff has nothing to do with COM. The objects containing
the data are C++ object. But when the mapping is complex, we create a COM
wrapper over those C++ object (ITKDataObject) and send them to a script
using ActiveX Scripting.

So, the problem is this. When I access "Employee.Name", I get the ID for the
Name property of Employee, but the Python COM caches it for ITKDataObject
(which, according to what Bill and you told me, is OK). If I access the name
property from another object (I mean another table), say department, then
the pythoncom tries to access the property usinf the ID generated on the fly
by Employee.

> Yes - VBScript does indeed re-query for the DISPID before each call -
> ie, it does not cache the items.  I would put money on the fact that
> this is your problem.
>
> I am quite confident that we do not have anything this fundamental
> broken.

So our code is not OK. You are right, I was wrong. I'm glad I didn't bet,
and, more important, I'm glad I have enough experience to fall in
finger-pointing mode. I was *quite* sure our code was OK, but I would have
been a fool if I had said I was 100% sure :)


> If it is the problem, you are clearly breaking the rules, and there is
> a good chance that later versions of VBScript, or indeed any other COM
> compatible language would also break.

Now, this is a potentially severe bug. For the time being, the integration
server is "locked" so all scripts must be written in VBScript, but it may
not be so in future version. Another thing that could go wrong is that there
is a new version of VBScript available. I haven't tried it yet, but I will:
maybe this version uses caching.

> ob = pythoncom.CoCreateInstance("Your.ProgID", None, pythoncom.CLSCTX_ALL,
pythoncom.IID_IDispatch)
> 
> dispid = ob.GetIDsOfNames("foo")
> print dispid
> 
> dispid2 = ob.GetIDsOfNames("foo")
> print dispid2
> 
> if dispid != dispid2:
>   print "Eeek"

That test would be useful (I would put it as an attachment to my bug report)
if I could make it work with the architecture I described earlier. I mean, I
must compare dispids from ITKDataObjects. 
I tried Employee.GetIDsOfNames("Name"), but I get:

  File "d:\Apps\Python\win32com\client\dynamic.py", line 400, in __getattr__
    raise AttributeError, "%s.%s" % (self._username_, attr)
AttributeError: <unknown>.GetIDsOfNames

I tried  a couple of things (like QueryInterface), but nothing worked. 

Anyway, thanks a lot for your help! 
I have endured this bug for six months without being able to point what the
problem was!

Thanks,
______________________________________________________
   Gaetan Corneau
   Software Developer (Quality Assurance Team)
   BaaN  Supply Chain Solutions  
   http://www.baan.com
   E-mail: Gaetan_Corneau at baan.com                  
   Tel: (418) 266-8252          
______________________________________________________
"Profanity is the one language all programmers know best"

 



From donn at oz.net  Sun Nov  7 13:53:53 1999
From: donn at oz.net (Donn Cave)
Date: 7 Nov 1999 18:53:53 GMT
Subject: open( "file","mode") difference between script and commandline?
References: <941942257@bigfoot.com>
Message-ID: <804hs1$jm2$0@216.39.151.169>

Quoth Michael Praschl <praschl at bigfoot.com>:
| im using python 1.5.2 for BeOS

Life can be sweet!

| and im having a problem with file opening. when i use 
|
| >>> f = open( "filename","w" ) 
|
| in the command line interface, this works perfectly ok. but when i use
| the same function from a script, i get the following error: 
|
| Traceback (innermost last): 
|   File "Signmail.py", line 19, in ? 
|     f = open( "/tmp/workfile", "w" ) 
| TypeError: illegal argument type for built-in operation 
|
| can anybody tell me why this is? and what i can do about it? 

Yes, but it's just lucky that you handed us the clue in the next
question.  You apparently start your program with
   from os import *

Now this imports an "open", the POSIX 1003.1 open() function whose
arguments are (string, int [, int])  You may want to use this function
some day, but what you should do the minute you read this message is
remove all "from module import *" from your programs and start fixing
up the module references.  The problem you just had is a terrific
example of the pernicious problems posed by this practice.

| the second problem i found while i was trying to get around this. i have
| a multiline string (linefeeds) in the variable reply, and want to do the
| following: 
|
| system( "echo >/tmp/pgpBeMail `reply`" )  #reply is within this "\"
| apostrphys (in case the newsreaders cant show them correctly) 
|
| in this case the file is empty after the call. i tried also turning to
| this "/" apostrophys and to normal ones, but then the file contains
| "reply" as expected, but not the contents of "reply" 

[Of course this will be os.system() once you've fixed that.]

Python string quoting isn't the art it is with some other more
shell-like interpreted languages.  You quote to get a string, and
then what's inside the quotes is pretty much the literal contents
of the string, there's no provision for evaluating variables in
there.  So the shell got your ` intact, and tried to execute "reply",
since that's its cue for a command expansion.

Now the string you need here should to include its own quotes around
the data you want to echo, to keep the shell from getting tangled up
interpreting that data - so you want the system() call to look like 
system( "echo > /tmp/pgpBeMail 'some reply data'" )  And note that
echo will add a newline to the end anyway.

So a couple of ways to generate this might be
    os.system("echo > /tmp/pgpBeMail '%s'" % reply)

    os.system("echo > /tmp/pgpBeMail '" + reply + "'")

	Donn Cave, donn at oz.net



From gmcm at hypernet.com  Thu Nov 25 08:16:12 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 25 Nov 1999 08:16:12 -0500
Subject: Python complaints
In-Reply-To: <19991125141028.G17130@longford.cs.monash.edu.au>
References: <383D1AA2.35E8F0EC@compaq.com>; from Greg Ewing on Fri, Nov 26, 1999 at 12:16:50AM +1300
Message-ID: <1268586725-8644924@hypernet.com>

Toby J Sargeant wrote:

> Just throwing this up in the air, and seeing what comes back...
> 
> class Acquire:
>    def __init__(self,lock):
>       self.lock=lock
>       lock.acquire()
> 
>    def __del__(self):
>       self.lock.release()
> 
> def function(mutex):
>   _a=Acquire(mutex)
>   # do stuff.
> 
> It relies on Acquire.__del__() being called as soon as _a goes
> out of scope, but that doesn't seem likely to change for the
> moment. (Of course doing this in JPython is a recipe for
> disaster... or at least extreme inefficiency)
> 
> It also has the advantage that you can exit the function in many
> ways (including by exception), and the lock will be released
> regardless.

Most unfortunately, exiting by exception will foul you up 
because the traceback will have a reference to the Acquire 
instance. You can clear it when you catch the exception, but 
since the consequences of forgetting to do that are both 
disastrous and mysterious, I think this idiom is best left to 
C++ (where I use it all the time).



- Gordon



From R.M.Everson at ex.ac.uk  Tue Nov  2 09:25:34 1999
From: R.M.Everson at ex.ac.uk (R.M.Everson)
Date: Tue, 2 Nov 1999 14:25:34 GMT
Subject: Drawing (directed) graphs
Message-ID: <ye1bt9das2p.fsf@ex.ac.uk>

Hello,

Does anyone know of Python code for drawing graphs of the
graph-theoretic type? Ideally, I'd like to be able to interactively
add nodes and edges etc and add data to structures associated with the
nodes and edges.

Many thanks
Richard Everson.

-- 
Exeter University     R.M.Everson at ex.ac.uk
http://www.dcs.ex.ac.uk/academics/reverson



From thamelry at vub.ac.be  Wed Nov 24 06:44:50 1999
From: thamelry at vub.ac.be (Thomas Hamelryck)
Date: 24 Nov 1999 11:44:50 GMT
Subject: Python complaints
References: <XFMail.991124113221.mikael@isy.liu.se>
Message-ID: <81gj3i$fe3$1@mach.vub.ac.be>

Mikael Olofsson (mikael at isy.liu.se) wrote:
[snip]
: I prefer to write readable, understandable code. To me, that means=20
: (among other things) indenting blocks. Hence, I already indent blocks
: in every programming language, if the language allows me to. So, block=20
: delimiters like begin...end or {} are simply redundant to me.
[snip]

I don't want to start the discussion about indentation again; it's no use
since indenting blocks is here to stay. The point is that if you want to 
write code that looks ugly or unreadable you will succeed in doing that,
indentation or not. In many cases, using a variable indentation (depending
on what construct you are using or the length of a line) leads to code
that is visuably much more appealing.

Cheers,

------------------------------------------------
Thomas Hamelryck//Free University Brussels (VUB)
Intitute for Molecular Biology// ULTR Department
Paardestraat 65//1640 Sint-Gensius-Rode//Belgium
------------------------------------------------



From claird at starbase.neosoft.com  Tue Nov 30 18:14:31 1999
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 30 Nov 1999 23:14:31 GMT
Subject: performance of tuple-less Python?
References: <1268226961-30285235@hypernet.com> <943917954snz@vision25.demon.co.uk> <slrn8471sd.m5h.nascheme@adler.dynodns.net> <slrn848jos.nnk.quinn@amatho.ugcs.caltech.edu>
Message-ID: <28AA06A136A68C1C.67A20460768084FC.DEC7E9AA6FF8E72D@lp.airnews.net>

In article <slrn848jos.nnk.quinn at amatho.ugcs.caltech.edu>,
Quinn Dunkan <quinn at amatho.ugcs.caltech.edu> wrote:
			.
			.
			.
>See any purely functional language, like haskell.  You can simulate mutability
>in haskell by passing monads around, but you might be surprised at how much
>you can get done with no side effects.
Indeed; people often *are* "surprised at how much
you can get done with no side-effects."
>
>And what was that party trick of changing 1 to equal 2 in forth?  Maybe that's
>more at the language level though, but I don't know much about forth (one of
>these days...).
There are several languages in which "1" can name
a variable, which might happen to have the value
2.  I can't think of one, though, in which the
integers are mutable in the sense of this thread.
-- 

Cameron Laird           http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX



From tismer at appliedbiometrics.com  Tue Nov 16 15:20:23 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Tue, 16 Nov 1999 21:20:23 +0100
Subject: list.without()?
References: <11A17AA2B9EAD111BCEA00A0C9B41793034AAEE7@molach.origin.ea.com>
Message-ID: <3831BC87.CD6ECCC4@appliedbiometrics.com>


"Stidolph, David" wrote:
> 
> so ...
> 
> def without(source, element):
>   while element in source:
>     source.remove(element)
> 
> Is that what you are looking for?  It would help if I had read the earlier
> posts!  :)

Semantically, yes. Proactically, I'm more keen on speed,
and i guess the single try catcher is more efficient than
to look up every thing twice. (Didn't measure it, but I'm quite sure).

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From mhammond at skippinet.com.au  Tue Nov 16 19:22:49 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 17 Nov 1999 00:22:49 GMT
Subject: Time object (COM)
References: <816010E2456BD111A48700805FBBE2EEFDF7D7@ex-quebec-u1.baan.com>
Message-ID: <tzmY3.5654$Xo.30075@news-server.bigpond.net.au>

Gaetan Corneau wrote in message
<816010E2456BD111A48700805FBBE2EEFDF7D7 at ex-quebec-u1.baan.com>...
>Hi,
>
>I have a COM object with a property that looks like this when I print it:
><time object at c5d430>
>
>What is it exactly? Where is it documented?

It is a time object :-)  It is documented in the "Win32 extensions" help
file under the PyTime object.

Mark.






From john.whittaker at alliedsignal.com  Tue Nov  9 17:23:21 1999
From: john.whittaker at alliedsignal.com (john whittaker)
Date: Tue, 09 Nov 1999 16:23:21 -0600
Subject: MS Word COM error w/ Borders property
Message-ID: <38289ED8.9BA2645A@alliedsignal.com>

I got the following error when I tried to use get at the bottom border
from a borders collection:
>>> borders = word.ActiveDocument.Tables[0].Borders
>>>borders[ -3 ]
Traceback (innermost last):
  File "<interactive input>", line 1, in ?
  File "C:\Program
Files\Python\win32com\gen_py\00020905-0000-0000-C000-000000000046x1033x8x0.py",
line 2834, in __getitem__
    return self._enum_.__getitem__(index)
  File "C:\Program Files\Python\win32com\client\util.py", line 37, in
__getitem__
    return self.__GetIndex(index)
  File "C:\Program Files\Python\win32com\client\util.py", line 56, in
__GetIndex
    raise IndexError, "list index out of range"
IndexError: list index out of range

The -3 is the value of the wdBorderBottom constant in Word Basic.  The
"borders" object is a COM border collection.  I'm not currently up on my
Python syntax, but I think that a value of -3 for any list index is
invalid in Python.  If that's so, how can I get at this border object
from Python?




From wfouche at mindspring.com  Sat Nov 13 23:35:22 1999
From: wfouche at mindspring.com (Werner Fouche)
Date: Sat, 13 Nov 1999 23:35:22 -0500
Subject: NT admin scripts with python
References: <000201bf2e34$26d93750$020a10ac@p350.gnaa.net>
Message-ID: <80le22$e7t$1@nntp2.atl.mindspring.net>

Geoff,

http://www.python.org/download/Contributed.html is a good place to start.
Since you are more interested in Win32 scripting issues, I suggest you get a
copy of the book "Python Programming on Win32" (see
http://www.python.org/psa/bookstore ) -- I have already ordered mine! It
really is amazing what you can do with Python on Win32.

Werner

"geoff nordli" <geoff at gnaa.net> wrote in message
news:000201bf2e34$26d93750$020a10ac at p350.gnaa.net...
> Excellent.  I will give it a go.
>
> Does anyone know if there is a reservoir of python scripts on the net.
Kind
> of
> like Perl's CPAN
>
> thanks,
>
> geoff
>
> -----Original Message-----
> From: python-list-admin at python.org [mailto:python-list-admin at python.org]
On
> Behalf Of Mark Hammond
> Sent: Saturday, November 13, 1999 3:39 PM
> To: python-list at python.org
> Subject: Re: NT admin scripts with python
>
> Yep - the win32net module is all you need!  Included in the win32all
> extensions...
>
> Mark.
>
> geoff nordli wrote in message
> <000101bf2e22$7060a190$020a10ac at p350.gnaa.net>...
> >Does python have a module that would allow a person to write
> >scripts on NT Server to lets say:  create users, map drives,
> >modify registry, create shares, and so on.
> >
> >thanks,
> >
> >Geoff Nordli
> >
> >
> >
>
>
>
> --
> http://www.python.org/mailman/listinfo/python-list
>
>
>





From m.faassen at vet.uu.nl  Sat Nov 13 08:58:48 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 13 Nov 1999 13:58:48 GMT
Subject: A Modest Proposal -- <wink>
References: <99110613580407.02970@quadra.teleo.net> <8072o6$m4i$1@newshost.accu.uu.nl> <y0jiu3b2357.fsf@vier.idi.ntnu.no>
Message-ID: <80jqqo$n3m$1@newshost.accu.uu.nl>

Magnus L. Hetland <mlh at vier.idi.ntnu.no> wrote:
> m.faassen at vet.uu.nl (Martijn Faassen) writes:

>> 
>> As you can see, XWTSNTSDL is fully consistent with your current proposals. 
>> It is therefore expected the major internet browsers will soon be able to
>> support these new standards.

> It is not consistent with XML... It would have to be <wink/>....

Ah, but XWTSNTSDL takes care of that, of course. :)

It defines a XPPMS (eXtensible PreProcessor Macro System), which allows
single <wink> tags. The XPPMS system transforms the source into valid XML.
After that the other subsystems of XWTSNTSDL take over.

I-can-come-up-with-more-acronyms-than-you-can-with-objections-ly yours, 

Martijn




From GabeN at captura.com  Mon Nov  1 14:02:17 1999
From: GabeN at captura.com (Gabe Newcomb)
Date: Mon, 1 Nov 1999 11:02:17 -0800 
Subject: Ultimate newbie question? How to execute DOS commands via Python.
Message-ID: <7309F1998D14D3119DD80008C7C9CA80631132@GIBRALTAR>

Hi all:
	My job here at work would be so much easier if  I could use Python
to run a series of DOS commands (I'd rather do this than write a batch file
for a number of reasons). Any pointers/examples on this would extremely
helpful!

Thanks in advance,

Gabe Newcomb



From harms at mbnet.mb.ca  Fri Nov  5 13:30:57 1999
From: harms at mbnet.mb.ca (Stefan Harms)
Date: Fri, 05 Nov 1999 13:30:57 -0500
Subject: Can Zope be used as a CVS server?
Message-ID: <harms-0511991330570001@192.168.1.35>

I would like to provide an arena for multiple workers to work on a project
together.

Could I use Zope for the functionality of a CVS server?
What about version controll?
What about the batch export/import functions of CVS - how would they be
executed in Zope?

(Maybe this is a stupid question, since you would think that the question
is answered by the simple fact that Zope source code is being kept on a CVS
server.  It does not hurt to ask, though.)

-- 
Stefan Harms 
LAMDI - Linux Anesthesia Modular Devices Interface
http://gasnet.med.yale.edu/lamdi/



From fredrik at pythonware.com  Wed Nov 10 05:03:44 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 10 Nov 1999 11:03:44 +0100
Subject: Alphanumerics
References: <19991110073503.13073.qmail@nwcst021.netaddress.usa.net>
Message-ID: <01d101bf2b62$e1477810$f29b12c2@secret.pythonware.com>

Alexis <alexisjuh at usa.net> wrote:
> I'm making some software development for a python program created in the US,
> well, it's pretty much just translating and adding some language specific
> modules for Norway, Sweden and Denmark...Problem is the special chars we have
> here, like ??? for norwegian..
> When I use a string everything works fine, but when I import fields from
> databases and they contain these chars I'll get escape seq instead...
> The chr(num) function doesn't work for these chars for some reason, and now
> I'm wondering if there's another way to get it right...

this is not really a problem in python, but the
environment can make it look much harder than
it is.   here are some basic facts, to help you
figure out what's going wrong.

    1. python's string type has no problem with
       non-ascii characters (as long as we're talking
       8-bit character sets, that is).  chr() and ord()
       works as you'd expect -- they convert between
       8-bit characters and unsigned integers.

    2. python has no default encoding -- you have
       to keep track of that yourself (or, easier, make
       sure you convert anything to, say, Latin 1).  if
       you don't know what encoding your data are
       using, figure that out before attempting to do
       something with that data.

    3. in cases where python uses "repr" to prepare
       strings for printing, it converts non-ascii character
       to /ddd representation, instead of sending them
       to the output device as is.

    4. the python interpreter uses "repr" to format
       strings when run in interactive mode, unless
       you use print:

            >>> s = "???"
            >>> s
            '\345\344\366'
            >>> print s
            ???

    5. while Windows usually use ISO Latin 1, MS-DOS
       (aka CONSOLE) WINDOWS DON'T.

            >>> s = "???"
            >>> s
            '\206\204\224'
            >>> print s
            ???

        (note the different output from repr).

okay, that's enough for now.  let us know if you still cannot
figure out where things go wrong in your program.

</F>

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




From aa8vb at yahoo.com  Wed Nov 17 08:58:57 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Wed, 17 Nov 1999 08:58:57 -0500
Subject: Line Benchmarks Tkinter vs. wxPython
In-Reply-To: <38318873.CE1B665E@bioreason.com>
References: <01bf2fa7$02992ca0$8a42c3d1@Cyberus.cyberus.ca> <slrn8318ec.9dg.bernhard@alpha1.csd.uwm.edu> <01bf2fda$9bf27f40$9c55c3d1@Cyberus.cyberus.ca> <19991116094209.A721671@vislab.epa.gov> <38318873.CE1B665E@bioreason.com>
Message-ID: <19991117085857.A807152@vislab.epa.gov>

Mitch Chapman:

 |You can also specify the option on the command line, e.g.
 |
 |build.py -b MAKE=gmake

Thanks for the tip.

 |> Then it failed finding wx.h.
 |
 |As you discovered, the fix is to unpack the wxPython tarball
 |under the utils directory of wxWindows.

Well, that wasn't a fix here. ;-)

   > pwd
   /scratch/ethyl/rhh/BUILD/wxPython/wxGTK

   > setenv WXWIN $cwd

   > cd utils/wxPython/src

   > python $WXWIN/utils/wxPython/distrib/build.py -b MAKE=gmake
   Makefile created:  Makefile
   Running: gmake -f Makefile
   CC -n32  -DSWIG_GLOBAL -DHAVE_CONFIG_H -DWXP_USE_THREAD -I. `wx-config --cflags` -I/home/rhh/bin/../software/python/include/python1.5 -I/home/rhh/bin/../software/python/include/python1.5 -I/scratch/ethyl/rhh/BUILD/wxPython/wxGTK/utils/wxPython/src -O -OPT:Olimit=0   -c helpers.cpp
   /bin/sh: wx-config:  not found
   cc-1005 CC: ERROR File = helpers.h, Line = 16
     The source file "wx/wx.h" is unavailable.

     #include <wx/wx.h>
                       ^

   1 catastrophic error detected in the compilation of "helpers.cpp".
   Compilation terminated.
   gmake: *** [helpers.o] Error 2


I'll table this effort for now.  Got some other things to take care of today.

-- 
Randall Hopper
aa8vb at yahoo.com



From kain at twilight.  Tue Nov  2 04:51:42 1999
From: kain at twilight. (Scott Barron)
Date: 2 Nov 1999 09:51:42 GMT
Subject: SWIG and function name question
Message-ID: <7vmc7e$8h$1@marcie.fissiontech.com>

Hey folks,

I downloaded and started to learn SWIG today and I ran into a little problem.
In the C code there is a structure like this:

struct __foo { ... };

When I import my SWIG created module using 'import mymodule' I have access 
to the struct's set/get operation via mymodule.__foo_set_member() but if I 
import my module using 'from mymodule import *' I have access to all of my
modules functions accept the two get/set functions that begin with __
(I'm using dir() to check this).  This isn't too much of a problem for me
but I'm just wondering what's going on.  Can anyone clear this up for me?

Thanks,
Scott




From reic0024 at ub.d.umn.edu  Tue Nov 16 12:03:21 1999
From: reic0024 at ub.d.umn.edu (Aaron J Reichow)
Date: Tue, 16 Nov 1999 11:03:21 -0600
Subject: Anyway to use file masks with os.listdir()?
In-Reply-To: <Pine.LNX.4.20.9911161645190.2604-100000@fep132.fep.ru>
References: <Pine.LNX.4.20.9911161645190.2604-100000@fep132.fep.ru>
Message-ID: <Pine.SO4.4.02.9911161102530.12652-100000@ub.d.umn.edu>

Thanks everyone for the replies!  When I go into to work today, I'll give
it a try. :)

(I knew it had to be simple!)

Aaron





From gregw at speakeasy.org  Tue Nov 16 02:42:39 1999
From: gregw at speakeasy.org (Greg Whelan)
Date: Mon, 15 Nov 1999 23:42:39 -0800
Subject: AIFC uncompression (where's module CL?)
Message-ID: <38310AEF.93CC9BA3@speakeasy.org>

I'm trying to use the aifc module to read a compressed AIFF-C file. The
documentation for the library states:

"Caveat: Some operations may only work under IRIX; these will raise
ImportError when attempting to import the cl module, which is only
available on IRIX."

Anyone know where I can find the CL module?



From andy at robanal.demon.co.uk  Sun Nov  7 17:04:48 1999
From: andy at robanal.demon.co.uk (Andy Robinson)
Date: Sun, 07 Nov 1999 22:04:48 GMT
Subject: Assorted Questions regarding the future of Python...
References: <002d01bf270e$238223a0$0501a8c0@bobcat>
Message-ID: <3827f770.36519111@news.demon.co.uk>

>And more to the point, VB was (and is), usually, "fast enough".  Who
>does huge matrix manipulation etc in VB anyway?  Lots of the focus the
>industry has on speed is a red herring IMO.
>
>Just like Python itself - really quite "slow" compared to these
>natively compiled languages - but in the entire time I have used it, I
>have had to use the profiler exactly twice.  In every other case, it
>has been more than "fast enough" (and the development speed far
>outweighs the difference anyway).
>
>Whats-a-milli-second-between-friends-ly,
>
>Mark.
>
The only thing I'd add is that you are skilled enough to use both
tools and make the right choice for the right job.  There are legions
of corporate "Application Programmers" (which I have come to realize
is a term of condescension used by "Systems Programmers") who only
know something like VB; occasionally they end up forced to do some
inappropriate data crunching in VB, or produce something which
thrashes the network or the database to death because they went for a
solution using only the tools they knew and not the best ones.  I was
guilty of this a few times in my VB days..

Andy




From fredrik at pythonware.com  Fri Nov 26 06:36:28 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 26 Nov 1999 12:36:28 +0100
Subject: Tkinter: closing app with active messagebox
References: <uogchfqlg.fsf@GARLIC.i-did-not-set--mail-host-address--so-shoot-me>
Message-ID: <003301bf3802$7beccdb0$f29b12c2@secret.pythonware.com>

<marten at btk.utu.fi> wrote:
> My question is: How do I force an active messagebox to close? Can this
> be done with the standard dialog boxes in tkMessageBox or do I have to
> write my own dialog boxes?

destroying the root window (call the "destroy"
method on the Tk instance) should work.

(or you can use os.exit()...)

hope this helps! (if not, please post some code).

</F>

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




From phd at phd.russ.ru  Sat Nov  6 08:00:51 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Sat, 6 Nov 1999 13:00:51 +0000 (GMT)
Subject: Get parameter by name
In-Reply-To: <19991106125145.EB3B17D071@anubis.uji.es>
Message-ID: <Pine.LNX.4.20.9911061300150.18974-100000@fep132.fep.ru>

On Sat, 6 Nov 1999, Gil Del Rosal David wrote:
> You can use the internal function vars(), e.g.:

   Yes! It works. Thank you!

Oleg.
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From a.eyre at optichrome.com  Tue Nov 30 09:01:19 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Tue, 30 Nov 1999 14:01:19 -0000
Subject: Python Path weirdness???
In-Reply-To: <81usqn$1lou$1@rtpnews.raleigh.ibm.com>
Message-ID: <001601bf3b3b$609ffc30$3acbd9c2@peridot.optichrome.com>

>  char* path ="D:\\POSSApps\\anotherposs app\\Embed Python\\some.exe";
>  Py_SetProgramName(path);

>From what I understand, you need:

char* path ="D:\\POSSApps\\anotherposs app\\Embed Python";
PyObject* pyPath = PyString_FromString(path);
PyObject* pythonPath = PySys_GetObject("path");
PyList_Append(pythonPath, pyPath);
Py_DECREF(pyPath);

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From hutton at chaffee.usyd.edu.chaffau  Tue Nov 23 19:36:41 1999
From: hutton at chaffee.usyd.edu.chaffau (Ross Hutton)
Date: 24 Nov 1999 00:36:41 GMT
Subject: Returning a variable reference
References: <slrn83jhr5.cl.hutton@el-047.ee.usyd.edu.au>
Message-ID: <slrn83mcos.j3.hutton@el-047.ee.usyd.edu.au>

Thanks to both Ben and Tom for their suggestions: The problem isn't
just one of indirection. It's being able to point inside (and modify) 
different lists with the same variable which I think is a NoNo in python
(most likely for safety reasons as pointed out by Ben). I should have been
more specific about the particular problem in my original post.

I think I've come up with a more appropriate solution (skip to end of post
if you're interested).

>Here's an idea; objects are lvalues. An excessively simple example
Problem with this is I need an lvalue that can point into either one
of two pre-existing lists.

>Uhm... Actually I think you're making trouble for yourself.  Use a
>list (or an array if you want bytewise access and can't afford to
>waste a little memory) for your memory area and return the index as
>your "address".  

I'm already doing this. I use separate lists for memory and
registers. Problem is that the same instruction needs to be able to
modify an element of either depending on an embedded 'rrr' code (eg:
A=111, B=000, (HL)=110).

At present I use a separate decoder routine that matches an opcode to
a valid z80 instruction and extracts the rrr code(s) if present. It
would be very nice for this routine to return an lvalue (that could
point to an element of either the register list or the memory list) for the
subsequent z80 instruction to use/modify. 

At present for every instruction that could be direct/indirect I
insert code to handle the indirect case as a special case (not very clean).

Another solution that I thought of is to reserve an area in the memory
list for all the registers to point to and make every register access
an indirect one.

Ahh, I think I've just seen the light. Rather than trying to pass an
lvalue up the chain I should be passing the rrr code down to lower
access functions (get(rrr) and set(rrr,value)) and have them do the
dirty work.

Many thanks again to Tom and Ben for their suggestions. You got me talking
to myself and that's probably not such a bad thing. 

It's quite interesting learning Python which is in many ways a lovely
language (problem is that I keep thinking in C). I was amazed when I
did some function pointer work without even looking in a manual. I
just assigned a function to a variable and called the variable as a
function. I just wrote code the way I thought it should work (and was
quite surprised when it did).

--
Ross (remove all chaff to reply)



From carroll at ebi.ac.uk  Tue Nov 30 10:31:10 1999
From: carroll at ebi.ac.uk (Mark Carroll)
Date: 30 Nov 1999 15:31:10 -0000
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk>
Message-ID: <820qju$q2s$1@pointy-haired-boss.ebi.ac.uk>

Thank you very much indeed for your replies! comp.lang.python seems
like a fairly nice place, and nobody mentioned any show-stoppers.
It's good to know what I might be getting into. I was very happy to
see a flame war fail to happen.

It was interesting to see people mention what might be coming in later
versions of Python (1.6, 2.0, whatever) - it'd be interesting to learn
a bit more about how and when the language might be changing slightly,
but I figure if I keep reading here I'll learn about that.

-- Mark



From greg.ewing at compaq.com  Mon Nov  8 08:55:08 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Tue, 09 Nov 1999 02:55:08 +1300
Subject: A Comparison of Python and Ruby Extension Interfaces (Re: 
 Comparison   between Python and "Ruby")
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp> <7vjtrc$hmm$1@newshost.accu.uu.nl> <38204614.2EAC6051@compaq.com> <5ld7trtak5.fsf@eric.cnri.reston.va.us>
Message-ID: <3826D63C.2731BB1F@compaq.com>

Guido van Rossum wrote:
> 
> Python used to have short names too, and in fact the current naming
> scheme was intruduced with great reluctance.  Unfortunately the
> realities of embedding Python in other apps required the current
> longer names

I know, I didn't mean that as a criticism of Python.
I understand the importance of keeping namespaces
under control!

In fact, I'm not very impressed with Ruby's naming
scheme, which seems rather inconsistent. Most Ruby-related
names start with "rb_", but some just start with "R",
and some don't have any unique prefix at all. 

Also it's not entirely clear what the rational is behind
the various uses of lowercase, uppercase, mixedcase,
embedded underscores, etc. There probably is one, but
it all looks a bit haphazard at first sight.

Greg



From skip at mojam.com  Mon Nov  1 12:03:55 1999
From: skip at mojam.com (Skip Montanaro)
Date: Mon, 1 Nov 1999 11:03:55 -0600 (CST)
Subject: Need Berkley DB!!
In-Reply-To: <381D5E84.98931682@bibsyst.no>
References: <381D5E84.98931682@bibsyst.no>
Message-ID: <14365.51195.136678.559903@dolphin.mojam.com>

    Thomas> After I spent most of my weekend trying to get Berkley DB to
    Thomas> work with Python on my Linux ( Suse 6.1) machine, I`ve now given
    Thomas> up completly.  I`ve tried to install DB.1.85. and DB.2.7.7,
    Thomas> neither worked, either crashing under installation ( which of
    Thomas> course, is not Python related ) or when I tried to test the
    Thomas> installation with the examples in the documentation ( the
    Thomas> wrapper for the 2.7.5-module listed at www.python.org, under
    Thomas> modules, databases ).

Berkeley DB is a strange beast.  It's available on a number of platforms
where it's used for things like password databases and sendmail mappings.
Unfortunately, this makes it difficult to upgrade without figuring out all
the places it's used and upgrading those tools as well.

(Not saying this is your problem, just speculating.)

there's-always-gdbm-ly y'rs,

Skip Montanaro | http://www.mojam.com/
skip at mojam.com | http://www.musi-cal.com/
847-971-7098   | Python: Programming the way Guido indented...



From mwh21 at cam.ac.uk  Mon Nov  1 16:29:34 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 01 Nov 1999 21:29:34 +0000
Subject: Produce .pyc without execution
References: <002001bf244e$8e187600$3acbd9c2@peridot.optichrome.com> <oq904ii2kj.fsf@titan.progiciels-bpi.ca>
Message-ID: <m37lk1dhoh.fsf@atrus.jesus.cam.ac.uk>

=?ISO-8859-1?Q?Fran=E7ois_Pinard?= <pinard at iro.umontreal.ca> writes:

[biggo snippo]

> Above, $(pkgdatadir) is an installation directory, so things are already in
> their final place, no need to move them further once there.  The scheme I
> used is acceptable when one wants to provide all sources installed with the
> compiled forms, but some people prefer to keep sources non-installed, just
> to spare disk space, say.  

I think installing .pyc's but not .py's is mischevious, if only
because meaningful tracebacks are nice.

> What would be ideal is a way for `compile_dir'
> to produce `.pyc' in the build hierarchy directly from the source hierarchy
> at `make all' time, and that `make install' just move these ready `.pyc'
> files into the installation directories.

It shouldn't be all that hard to adapt Lib/py_compile.py &
Lib/compileall.py as appropriate to effect this.

Regards,
Michael



From charles.choi at computer.org  Tue Nov 23 12:54:57 1999
From: charles.choi at computer.org (Charles Y. Choi)
Date: Tue, 23 Nov 1999 09:54:57 -0800
Subject: How to perform syntax check in Python
References: <382A2384.CC37FBA8@hp.com> <00bf01bf2c1d$2a927be0$f29b12c2@secret.pythonware.com> <3838CE1A.C7BE0530@horvath.com> <slrn83hl31.p19.wtanksle@hawking.armored.net> <38398D45.DFA52EFD@horvath.com>
Message-ID: <383AD4F1.740DFD7E@computer.org>


Bob Horvath wrote:
> 
> Having said that, and being in basic agreement with your statement, that doesn't
> answer the original question.  Are thre any coverage analyzers for Python code?

Look for trace.py by Andrew Dalke.


-Charles



From volumeregeling at hotmail.com  Tue Nov 30 09:51:54 1999
From: volumeregeling at hotmail.com (volumeregeling)
Date: Tue, 30 Nov 1999 14:51:54 GMT
Subject: COMobject problem
Message-ID: <3843e3f3.18656326@122.0.0.250>

Hi,

i have a problem regarding COM objects, i invoke the object through
the following statement :

Editor= win32com.client.Dispatch("Comeditor.Classeditor")

Now i have the problem that when i close the main program , the
COMobject will still be alive, i want to close all the comobjects
opened from within python to be closed when the main program closes.

I have no idea on how to do that, maybe you do , then i will be very
gratefull.



thanks



From mikael at isy.liu.se  Fri Nov 19 05:00:42 1999
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Fri, 19 Nov 1999 11:00:42 +0100 (MET)
Subject: Beginning programming with Python
In-Reply-To: <Pine.LNX.4.10.9911191043080.2313-100000@essle.valdyas.org>
Message-ID: <XFMail.991119110042.mikael@isy.liu.se>

On 19-Nov-99 Boudewijn Rempt wrote:
 >  On Fri, 19 Nov 1999, Fredrik Lundh wrote:
 >  
 > > Boudewijn Rempt <boud at rempt.xs4all.nl> wrote:
 > > > I tend to forget the colons after function and class definitions and
 > > > everywhere else and the double == in if's - mainly because, for the life
 > > > of Brian, I can't see why they are there. It seems to me the language
 > > > would be just as easily parseable and readable without this bit of
 > > > interpunction.
 > > 
 > > guess you weren't part of the CWI design team -- they
 > > did test this on humans, you know, and discovered that
 > > the language wasn't as easily parseable and readable
 > > without the colons...
 > > 
 >  
 >  You needn't guess about that, I guess ;-). I didn't know there
 >  was a design team for Python, I though it was Guido's invention.
 >  But I still wonder about the double == - if I can't assign in
 >  an if, why insist on ==?

Consider the following:

>>> a=b==c
>>> if a:
...   print "foo"

Then compare that to the following:

>>> if b==c:
...   print "foo"

As I understand your thought, you want the following to be used:

>>> if b=c:
...   print "foo"

I would vomit if I had to write the test b==c as b=c sometimes, just 
because there is no ambiguity in those cases, which is what you are 
suggesting we should do.

My-keyboard-would-become-messy-ly y'rs

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    19-Nov-99
Time:    10:50:07

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



From mwh21 at cam.ac.uk  Sat Nov  6 19:10:24 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 07 Nov 1999 00:10:24 +0000
Subject: Job skills [was Re: making the Python case to my dept.]
References: <1270209151-4344374@hypernet.com>
Message-ID: <m3n1srnoun.fsf@atrus.jesus.cam.ac.uk>

"Gordon McMillan" <gmcm at hypernet.com> writes:

> Michael Hudson asks:
> 
> > "Thomas A. Bryan" <tbryan at python.net> writes:
> > 
> > > Do math majors or engineers learn their critical job skill's in
> > > their first classes?  
> > 
> > What are mathematicians' critical job skills? I'm curious, being
> > one, and not having being taught anything resembling a `critical
> > job skill' yet, and happy about it.
> 
> The ability to work long and hard on entirely meaningless 
> problems.
> 

Fair comment.

Cheers,
Michael



From nbecker at fred.net  Tue Nov  9 16:45:36 1999
From: nbecker at fred.net (nbecker at fred.net)
Date: 09 Nov 1999 16:45:36 -0500
Subject: what is a function object?
References: <809p0a$8nf$1@nnrp1.deja.com> <Ec0W3.24420$Ec1.274008@news1.frmt1.sfba.home.com>
Message-ID: <x8866zbuynz.fsf@rpppc2.hns.com>

>>>>> "claudius" == claudius  <claudius at catlover.com> writes:


    >> * What does it mean to say that function objects are first-class objects
    >> in the python environment?

    claudius> Good question...I have no idea.

I think it means you can assign a function object to a variable?



From ionel at psy.uva.nl  Fri Nov 12 13:01:55 1999
From: ionel at psy.uva.nl (Ionel Simionescu)
Date: Fri, 12 Nov 1999 19:01:55 +0100
Subject: docs
References: <80epm7$vbo$1@nnrp1.deja.com> <14378.61648.633649.431232@weyr.cnri.reston.va.us>
Message-ID: <80hkju$okf@mail.psy.uva.nl>

Fred L. Drake, Jr. <fdrake at acm.org> wrote in message
news:14378.61648.633649.431232 at weyr.cnri.reston.va.us...
|
| mehta at mama.indstate.edu writes:
|  > I have noticed for the local docs of python in IE5 on winnt pages
|  > randomly donot get rendered but are left as raw html in browser as I go
|  > back and forth using browser buttons.  has any one experienced same
|
|   I have never seen this; if anyone can reproduce this and help
| isolate the cause I'd appreciate hearing about it.

I have seen it before and I have just checked it again.

Just open a doc,
say C:\Python\Doc\tut\tut.html,
with Internet Explorer 5.00.2314.1003,
on Windows NT 4.0 SP4,
and refresh it (F5).


ionel







From hhh14021978 at my-deja.com  Wed Nov 17 14:39:48 1999
From: hhh14021978 at my-deja.com (hhh14021978 at my-deja.com)
Date: Wed, 17 Nov 1999 19:39:48 GMT
Subject: Tk title bar
Message-ID: <80v0a3$aei$1@nnrp1.deja.com>

Hi,

I'm new with python-tk. I want to change the icon
in the title bar. Is there any way to do that?

Thanks,

Hans


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



From g_umberger at mail.utexas.edu  Wed Nov 10 14:33:59 1999
From: g_umberger at mail.utexas.edu (George Umberger)
Date: Wed, 10 Nov 1999 19:33:59 GMT
Subject: a new parallel meta search engine 
Message-ID: <HMjW3.21383$l4.153568@typhoon2.austin.rr.com>

Hi everyone,
Sorry for cross posting.

I've found a new parallel meta search engine that searches
over 18 search engines in one click.

Please check it out for yourself.

http://meta360.com

thanks

George.




From jam at quark.emich.edu  Wed Nov 24 07:32:34 1999
From: jam at quark.emich.edu (Jeff)
Date: Wed, 24 Nov 1999 07:32:34 -0500
Subject: How to capture stdout in an exec'ed string?
In-Reply-To: <80v7j4$g0n$1@nnrp1.deja.com>
References: <80v7j4$g0n$1@nnrp1.deja.com>
Message-ID: <19991124073234.B25371@quark.emich.edu>

On Wed, Nov 17, 1999 at 09:44:11PM +0000, Preston Landers wrote:
> Hello all.
[..snipped..]
> 
> At no time do I want the old program's output to be visible to the
> user.  Therefore I must find some way trap the print statements.
> (Going through the old code and changing all the prints to a custom
> function call is not really an option at this point.)
> 
> In the subject line of this msg I said an "exec'ed statement" but it
> not really need be.  I can just instantiate the old program and call
> the methods I need.  Something like this:
> 
> try:
>    cruft = my_ancient_class()
>    cruft.__stdout__ == My_STDOUT()
>    cruft.do_your_thang()
> except SystemExit, msg:
>    process(msg)
> 
> I guess my question is how to accomplish the 3rd line of my code.  What
> kind of class does My_STDOUT() need to be?  Something mimicing a file
> handle?  How to actually "attach" it?  My guess is that I will have to
> eval my_ancient_class in a special namespace?
> 

how about just calling 'open' and assigning it to 'sys.stdout'? your
assumption is correct-- stdout is nothing more than a standard file handle
(under UNIX anyway).

you could try something like this (untested):

sys.stdout = open("/dev/null", "w")
cruft = my_ancient_class()
try:
 cruft.do_your_thang()
except SystemExit, msg:
 process(msg)

you may also want to keep track of the 'old' stdout file handle so it can be
restored later.

hope that helps.

regards,
J
-- 
|| visit gfd <http://quark.emich.edu/>
|| psa member #293 <http://www.python.org/> 
|| New Image Systems & Services, Inc. <http://www.newimage.com/>



From rumjuggler at home.com  Mon Nov  1 19:32:25 1999
From: rumjuggler at home.com (Ben Wolfson)
Date: Tue, 02 Nov 1999 00:32:25 GMT
Subject: newbie intro recs?
Message-ID: <381E3205.BEFDE12C@home.com>

Can someone recommend a good introduction to Python, for a very novice
programmer?  (the only background I have is c++ from ap comp sci at high
school, and what I've gleaned from the C++ Programming Language, so I
can't pick things up as easily as I'd like).

Thanks!

-- 
Barnabas T. Rumjuggler

Graybags: "Irony = metally" (U.S. dictionary)
Me: And here I thought it was the practice of removing wrinkles from
clothes.
Graybags: Removing wrinklies from clothes is euthanasia.
 -- in alt.possessive.its.has.no.apostrophe



From bjorn at roguewave.com  Mon Nov 29 16:03:36 1999
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Mon, 29 Nov 1999 14:03:36 -0700
Subject: Python bugs
Message-ID: <51C9F3C392D0D111A15600A0C99E64A9AF569C@localhost>

>     Python 1.5.2 (#0, Apr 13 1999, 10:51:12) [MSC 32 bit 
> (Intel)] on win32
>     Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>     >>> import os
>     >>> {}.update(os.environ)
>     Traceback (innermost last):
>       File "<stdin>", line 1, in ?
>     TypeError: dictionary, instance
>     >>> import shelve
>     >>> s = shelve.open('foo')
>     >>> s.get('bar', 'baz')
>     Traceback (innermost last):
>       File "<stdin>", line 1, in ?
>     AttributeError: get
>     >>> 
> 
> I believe os.environ is a class to allow constructs like
> 
>     os.environ["FOO"] = "PYTHON"
> 
> to work like
> 
>     os.putenv("FOO", "PYTHON")
> 

Which is a good thing, but shouldn't preclude it from being used as a
dictionary in the update statement (although the fix probably have to be to
Dict.update rahter than os.environ).

> with regard to setting or changing the values of environment 
> variables.
> 
> The lack of a get attribute in shelve objects seems to have 
> been rectified
> since the 1.5.2 release:
> 
>     >>> db = shelve.BsdDbShelf (bsddb.btopen("/var/tmp/foo2", "c"))
>     >>> db["1"] = 1
>     >>> db.keys()
>     ['1']
>     >>> db.get("1", 0)
>     1
>     >>> db.get("2", 0)
>     0                        

Cool (do you know if you can do db.items() also? -- found that one five
minutes ago...)

-bjorn



From Stuarty at excite.co.uk  Fri Nov 26 10:06:43 1999
From: Stuarty at excite.co.uk (Stuarty at excite.co.uk)
Date: 26 Nov 1999 15:06:43 GMT
Subject: NEWBIE
Message-ID: <81m7m3$82$1@news.qub.ac.uk>

I have just recently started using Python, after the first day I was further 
on than I am after a semester of C programming(were at arrays...).
tutorial by Giudo van Rossum was very useful, but I attempted printing out the 
reference library..there are now less considerably less trees in the world.

Stuart McFadden.
Stuarty at excite.co.uk




From curtis_c at lxe.com  Thu Nov 18 18:24:25 1999
From: curtis_c at lxe.com (Curtis, Craig M.)
Date: Thu, 18 Nov 1999 18:24:25 -0500
Subject: Python in DOS
Message-ID: <DB6390A9770ED211B2440000F807FC7B01039B8A@nt_exlxe.lxe.com>

Does anyone know the "approved" way to get rid of the following message:

Could not find platform dependent libraries <exec_prefix>
Consider setting $PYTHONHOME to <prefix>[:<exec_prefix>]
'import site' failed; use -v for traceback
Python 1.5.2 (#6, Nov 18 1999, 17:19:05)  [GCC 2.8.1] on ms-dos7
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam


I've tried everything I can think of to set PYTHONHOME and PYTHONPATH and
python ALWAYS starts up this way.

Thanks,
Craig Curtis



From nemeth at iqsoft.hu  Sat Nov  6 05:14:43 1999
From: nemeth at iqsoft.hu (Németh Miklós)
Date: Sat, 06 Nov 1999 11:14:43 +0100
Subject: making the Python case to my dept.
References: <9911051740130D.02530@william.jinsa.org> <7GOU3.272$Ae3.194028@news-dal.corridex.com>
Message-ID: <3823FF93.A5B3C4C7@iqsoft.hu>

Evan Simpson wrote:

> William Ahern <wahern at jinsa.org> wrote in message
> news:9911051740130D.02530 at william.jinsa.org...
> > I have never used Python, but I'm currently attempting to persuade a
> department
> > at my university (Geroge Washington) to use Python as an intro language. I
> have
> > read that it was good for this. From what I hear, the current options on
> the
> > table are Delphi and Java. I don't know much about Delphi, but I doubt
> Java is
> > any good for intro classes.
>
> Well, I don't know much about Java, but I've been a professional Delphi
> programmer for about five years now.  Up until a year and a half ago, I'd
> have said that Delphi would be fine for your purposes.  Then I learned
> Python :-)

Well, I don't know much about Delphi, but I've been a professional Java
(and
C++) programmer and project manager for about three years now. Up until
a year
ago, I'd have said that Java would be fine for your purpose. Then seeing
several terrible Java applications I started hesitating: is Java really
so good
as the herds of marketing peaple of big corporations say? Half a year
ago I met
Python, and now I know Python is the "king" language. Java is too
complex, not
as complex as C++ but, to complext to learn it easily. Python is much
more
easier, it supports an unprecedented dynamic OO technology not found in
Java or
C++.

NM





From PattyThomas at news.vex.net  Sat Nov  6 17:03:43 1999
From: PattyThomas at news.vex.net (Patty Thomas)
Date: 6 Nov 1999 22:03:43 GMT
Subject: python: something completely different
References: <3820C0B8.62B2@iowegian.com> <38249ED2.BB7EA1DE@znet.com>
Message-ID: <8028jv$7cf$1@hub.org>

On Sat, 06 Nov 1999 13:34:10 -0800, Robert Cain <arcane at znet.com> wrote:
>Back to the subject at hand (Python in the DSP universe) has anyone
>investigated the possibility of conversion, compilation or whatever of
>MatLab to Python?  I'm naive WRT to both at this point because 1) MatLab

I know nothing about MatLab, but a quick search here
http://www.vex.net/~x/parnassus/apyllo.py?so=d&find=matlab reveals that it
would seem Travis Oliphant has some MatLab like tools (or functions at least)
which could be useful until someone 'converts, compiles, or whatever', <grin>
matlab to python here... http://oliphant.netpedia.net/




From phil at ricochet.net  Mon Nov 29 02:37:32 1999
From: phil at ricochet.net (phil)
Date: Mon, 29 Nov 1999 07:37:32 GMT
Subject: PyArg_ParseTuple() and oprtional remaining argument lists
Message-ID: <38422744.174116386@news.ricochet.net>

Hi,

i've developed an extension .dll in C for a Window's based system and
it works really well. However, i wanted to add a function that took a
second optional argument so wrote the following:

...blah...
	if (PyArg_ParseTuple(args,"s|s",&name,&show)) {
		// happy things.

and it seems to return 0 if only one parameter is supplied, yet is
happy with two. It's not a major inconvenience as i can just document
it as requiring two parameters, but i'd like to know what i'm doing
wrong here. I also have trouble using PyArg_ParseTuple(...) with a
single item, e.g.
	if (PyArg_Parse(args,"i",&index) && (index >= 0)) {
is ok, but
	if (PyArg_ParseTuple(args,"i",&index) && (index >= 0)) {
fails. Again not a major problem, i'm more concerned that i'm doing
something wrong that will bite me later (this was cut and pasted from
my code, so the syntax may be a little suspect...)

The documentation shows examples of the above working correctly. I'm
using 1.52. I'm sure it's something silly.

Gracias  Mi Llama es su Llama.

phil.

The world is divided into two sorts of people: those that think the
world is divided into two sorts of people and those that don't.



From wware-nospam at world.std.com  Sun Nov 28 23:31:53 1999
From: wware-nospam at world.std.com (Will Ware)
Date: Mon, 29 Nov 1999 04:31:53 GMT
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <81bmmi$279$1@nnrp1.deja.com> <3841EE77.6B5D6AC9@wjk.mv.com>
Message-ID: <FLxz96.IEI@world.std.com>

William J. King (wjk at wjk.mv.com) wrote:
: 4. conversion of integer to a string ( which may exist but havn't found
: yet)

Try this:
	str(123)
or:
	hex(123)

: 2. would like something like  'increment'/'decrement' operators
:         either in form:
:                     "x++"  or "incr x" vice "x = x + 1 " but its ok

I find myself grumbling about having to type "x = x + 1". The really
clean thing to do, given that integers are objects, would be to define
increment and decrement methods, so you'd type something like "i.incr()".
Then you'd want to go around looking for other objects for which
it made sense to increment and decrement. Long ints and floats can do
what integers do. Sequences (lists, tuples, strings, xranges) might
grow and shrink in length. No good idea what to do for dictionaries.
-- 
 - - - - - - - - - - - - - - - - - - - - - - - -
Resistance is futile. Capacitance is efficacious.
Will Ware	email:    wware @ world.std.com



From piers at cs.su.oz.au  Sun Nov 28 23:50:52 1999
From: piers at cs.su.oz.au (Piers Lauder)
Date: Mon, 29 Nov 1999 15:50:52 +1100
Subject: threading and timeouts
Message-ID: <943851057.889.181768497@cs.usyd.edu.au>

The responses to my request on how to timeout threads yielded one very
useful module from Phil Mayes :-

	timeout_socket.py

This module defines a `timeout_socket' class for connections that can
potentially cause a server to hang.   By sub-classing and inheriting
from standard modules like smtplib, one can override the socket code
and use the original module without modification. See the examples
provided in the __doc__ string for more details.


With Phil's encouragement, I've modified it slightly so that it works on UN*X
systems, and have submitted it to the Vaults of Parnassus.

You can access it directly here:

	http://www.cs.su.oz.au/~piers/python/timeout_socket.py

Thanks Phil!







From bkenniso at worldnet.att.net  Tue Nov 30 08:35:00 1999
From: bkenniso at worldnet.att.net (Brian Kennison)
Date: 30 Nov 1999 13:35:00 GMT
Subject: Pthreads
References: <81v2an$n6o$1@bgtnsc01.worldnet.att.net> <81vhaq$cri$1@towncrier.cc.monash.edu.au>
Message-ID: <820jq4$cbt$1@bgtnsc02.worldnet.att.net>

I tried this but with the same results. Obviously it doesn't like the 
file it's in (parse error) but I don't know enough as to what it's 
complaining about. 

Thanks for you help though,

Brian

-----------------------------------------------
Jonathan Giddy (jon at rdt.monash.edu.au) wrote:
: bkenniso at worldnet.att.net (Brian Kennison) writes:

: >I'm having trouble compiling Python 1.5.2. with the FSU pthread
: >library. I working on MachTen and get the following error:

: >gcc -g -O2 -I./../Include -I.. -DHAVE_CONFIG_H   -c thread.c -o thread.o
: >In file included from thread.c:150:
: >thread_pthread.h:113: parse error before `pthread_cond_t'
: >thread_pthread.h:113: warning: no semicolon at end of struct or union
: >thread_pthread.h:114: warning: data definition has no type or storage class
: >thread_pthread.h:115: warning: data definition has no type or storage class
: >thread_pthread.h: In function `PyThread_start_new_thread':       

: I'm not sure this is the same problem, but I've seen similar messages under
: different circumstances (freezing with extension modules on Solaris).

: The problem in that case was that the file Include/thread.h in the Python 
: source clashed with the system thread.h file.  Try deleting Include/thread.h 
: and recompiling.  (In 1.5.2, thread.h is unused by Python, although it may 
: be used by some extensions).

: Jon.



From lordxiphias at altern.org  Wed Nov  3 05:31:38 1999
From: lordxiphias at altern.org (lordxiphias at altern.org)
Date: 3 Nov 1999 10:31:38 GMT
Subject: LX.MP3z: NO PORN, NO POP-UPS, NO BROKEN LINKS (Just MP3z!!)
Message-ID: <7vp2ua$28jm1416@michener.tea.state.tx.us>

LX.MP3z
***********
New upcoming MP3-site without porn, pop-ups or porn!!
With own private server => NO BROKEN LINKS
ACCESS FOR EVERYONE!!!
Free downloads...

Check It Out:
http://dowload.at/LX.MP3z




From wtanksle at hawking.armored.net  Wed Nov  3 08:27:51 1999
From: wtanksle at hawking.armored.net (William Tanksley)
Date: 3 Nov 1999 13:27:51 GMT
Subject: How can I make my assertions smarter?
References: <381D7087.BF8510E5@webone.com.au>
Message-ID: <slrn8216ve.j7h.wtanksle@hawking.armored.net>

On Mon, 01 Nov 1999 21:50:47 +1100, Stuart Hungerford wrote:
>Hi all,

>I'm using some simple functions to implement crude precondition,
>postcondition and invariant checking in Python classes.

Cool.  Unfortunately, to get anything like Eiffel, you'd have to modify
the language (add a keyword), because such things require a statement, not
an expression.  That's not impossible, although it might be more work than
you're willing to do, and it would certainly not become part of Python 1.x
(to say the least).

>Is there some neat way to wrap these tests up so I can print any
>code expression that fails, e.g:

>     Precondition failed foo.py line 42: spoon.shape == 'bent'
>which would appear at line 42 of foo.py as
>     pre(spoon.shape == 'bent')

assert spoon, "There is no spoon"

>I can see how to use environment variables or global variables to
>selectively enable precondtions etc, although any hints for doing that
>would be appreciated too.

Preconditions can also be implemented via aspects.  Aspect Oriented
Programming is _really_ cool, and with the author of JPython helping with
AspectJ I suspect that Python will someday handle them.

Aspects are really cool -- like a semantic preprocessor on steroids.

>Stu

-- 
-William "Billy" Tanksley



From fdrake at acm.org  Fri Nov 12 09:58:58 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Fri, 12 Nov 1999 09:58:58 -0500 (EST)
Subject: Tkinter and after
In-Reply-To: <382C2854.E9494EFE@lri.fr>
References: <382C2854.E9494EFE@lri.fr>
Message-ID: <14380.11058.151680.518510@weyr.cnri.reston.va.us>

Stephane Conversy writes:
 > It is said in Tkinter doc that after() is like a call to time.sleep()
 > wich blocks Tkinter.

  I'm not sure what docs you're referring to (presumably the
eff-bot's), but after() does not block Tk.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From ejohnson at netcom.com  Thu Nov 25 13:07:45 1999
From: ejohnson at netcom.com (Eric Johnson)
Date: 25 Nov 1999 18:07:45 GMT
Subject: Cons (replacement for make) in Python?
References: <Bbe%3.8183$cz3.292454@nnrp1.uunet.ca>
Message-ID: <81jtth$aa2$1@nntp8.atl.mindspring.net>

"Greg Wilson" <gvwilson at nevex.com> writes:

> Hi.  Has anyone built a tool like Cons (a pure-Perl replacement for
> make) in Python?  If so, I'd welcome a pointer.

I've written such a beast, but have yet to release it.  I'm still
working on its web site and plan to unveil it shortly.  It has
multi-threaded support so it supports parallel building in a very
natural way.  It should run on any OS that supports Python 1.5.2.

If you're interested, please let me know.

-Eric Johnson



From harms at mbnet.mb.ca  Sat Nov  6 09:47:26 1999
From: harms at mbnet.mb.ca (Stefan Harms)
Date: Sat, 06 Nov 1999 09:47:26 -0500
Subject: Can Zope be used as a CVS server?
References: <harms-0511991330570001@192.168.1.35> <14371.11961.82152.303247@dolphin.mojam.com>
Message-ID: <harms-0611990947260001@192.168.1.35>

In article <14371.11961.82152.303247 at dolphin.mojam.com>, skip at mojam.com (Skip Montanaro) wrote:

>     Stefan> Could I use Zope for the functionality of a CVS server?
> 
> Dunno, and it's not obvious (to me) why you'd want to do that.  Could you
> motivate your question by providing some rationale for not using CVS itself?
> 
> Skip Montanaro | http://www.mojam.com/
> skip at mojam.com | http://www.musi-cal.com/
> 847-971-7098   | Python: Programming the way Guido indented...


The only reason for using Zope as a CVS server, is that somebody I know finds it difficult to get CVS going as a server, and that the potential participants are not familiar with CVS, but they all have browsers...
and therefore can learn to use Zope.

Thanks for the reply

-- 
Stefan Harms 
LAMDI - Linux Anesthesia Modular Devices Interface
http://gasnet.med.yale.edu/lamdi/



From gerrit.holl at pobox.com  Thu Nov 18 06:08:59 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Thu, 18 Nov 1999 12:08:59 +0100
Subject: UserDict, UserList... UserFile?
Message-ID: <19991118120859.A2974@optiplex.palga.uucp>

Hello,

why isn't there a UserFile class, while there is a UserList and a UserDict?
And what about other builtin-types for which this is possible?

Otherwise, I'll write it myself; it shouldn't be too hard!

-- 
"I'd crawl over an acre of 'Visual This++' and 'Integrated Development
That' to get to gcc, Emacs, and gdb.  Thank you."
(By Vance Petree, Virginia Power)



From hinsen at cnrs-orleans.fr  Mon Nov 29 10:27:41 1999
From: hinsen at cnrs-orleans.fr (Konrad Hinsen)
Date: 29 Nov 1999 16:27:41 +0100
Subject: Numeric write-in place facility
References: <81tovm$oe0@mail.psy.uva.nl>
Message-ID: <m3bt8dz5aq.fsf@chinon.cnrs-orleans.fr>

"Ionel Simionescu" <ionel at psy.uva.nl> writes:

> Playing around I see now that other ufuncs allow a write-in target
> specification as their last argument. For example,
> 
> Numeric.sqrt(a,b)
> 
> will put the result in b (if its dimension and type are Ok).
> 
> The docs do not tell much about this.
> Can anyone confirm/clarify/comment ?

I can confirm that the feature exists and that it is well hidden ;-)

This was implemented to permit the fast evaluation of complicated
formulas without the allocation of temporary arrays for each
intermediate. Using this approach means more work and less readable
code, but also faster execution and lower memory needs.
-- 
-------------------------------------------------------------------------------
Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.55.69
Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais
-------------------------------------------------------------------------------



From x at x.x  Sat Nov 13 19:16:25 1999
From: x at x.x (Eric Jacobs)
Date: Sat, 13 Nov 1999 19:16:25 -0500
Subject: method overriding trick
References: <14378.62498.649973.150075@bitdiddle.cnri.reston.va.us> <lK9X3.696$An3.9333@newsfeed.avtel.net>
Message-ID: <382DFF59.51F186A6@x.x>

Phil Mayes wrote:
> 
> I use a method that doesn't remove the clunkiness, but localises the
> hierarchy to one line:
> 
> _Parent = basedlg.BaseDlg
> class AddressDlg(_Parent):
>     def __init__(self, param):
>         self.local = param
>         _Parent.__init__(self, 2345)
> 
> By using the same name "_Parent" in all modules, less thinking is needed,
> and cut'n'paste coding (what, me?) works better too.

You wouldn't be able to reassign the _Parent variable to define another
class elsewhere in the same module, though, because the functions defined
this way will always look up the _Parent from the module namespace at
execution time. A variant that removes that restriction is:

> _Parent = basedlg.BaseDlg
> class AddressDlg(_Parent):
>     def __init__(self, param, _Parent=_Parent):

The default args are evaluated at function definition time, which means
that the variable has the value you expect. Now you have true cut'n'paste
coding!



From kcazabon at home.com  Sun Nov 14 23:43:55 1999
From: kcazabon at home.com (Kevin Cazabon)
Date: Mon, 15 Nov 1999 04:43:55 GMT
Subject: win32api... another question:  GetEnvironmentVariable("PATH")
Message-ID: <fcMX3.65012$Rx2.773564@news1.rdc1.bc.home.com>

I can't seem to get the win32api.GetEnvironmentVariable() function to work
properly... am I not using it correctly by the following?

path = win32api.GetEnvironmentVariable("PATH")

or

path = win32api.GetEnvironmentVariable("%%PATH%%")

BOTH of the above return None to path... what am I missing?

And lastly, is there an equivalent "SetEnvironmentVariable"?  There is no
mention in the docs, but maybe there's a way to modify the Windows PATH on
the fly, besides a os.system() command (which isn't really correct either).

Thanks...

Kevin.
kcazabon at home.com





From jim at interet.com  Wed Nov  3 15:22:31 1999
From: jim at interet.com (James C. Ahlstrom)
Date: Wed, 03 Nov 1999 15:22:31 -0500
Subject: Newbie question : Info on Python under DOS/Windows
References: <381ec4c4.7451128@212.225.36.99>
Message-ID: <38209987.2A7005C3@interet.com>

Endgamer wrote:
> hold the date on reboot after the millenium.
> My question is: I've found it hard to locate resources or  information
> on Python under DOS/Windows - are there any special groups, sites or

Python on DOS/Win95 works the same as on Unix.  Just
start at www.python.org.

Jim Ahlstrom



From ivnowa at hvision.nl  Wed Nov 24 18:14:06 1999
From: ivnowa at hvision.nl (Hans Nowak)
Date: Thu, 25 Nov 1999 00:14:06 +0100
Subject: mysql and DATE format
In-Reply-To: <81hhm6$ba2$1@nnrp1.deja.com>
Message-ID: <199911242313.AAA06063@axil.hvision.nl>

On 24 Nov 99, sp00fD wrote:

> I've currently got some perl code that executes a sql statment which
> looks like:
> 
> SELECT id, subject, author, text, DATE_FORMAT(date, \"%W, %d %b %Y\"),
> DATE_FORMAT(date, \"%h:%i %p\") FROM Table WHERE thread=18 ORDER by date
> desc
> 
> When trying to use that in python like this:
> """SELECT .. DATE... FROM %s WHERE thread=%d""" % (table, id)
> 
> it tries to expand the DATE %W.. variables.
> I've also tried it as
> 
> sql = 'SELECT id, subject, author, text, '
> sql = sql + 'DATE_FORMAT(date, \"%W, %d %b %Y\"), '
> sql = sql + 'DATE_FORMAT(date, \"%h:%i %p\") FROM' + table
> sql = sql + 'WHERE thread=' + id + 'ORD...'
> 
> which gives me a TypeError: illegal argument type for built-in
> operation.  Why?  How can I do this properly?

Do I understand correctly that you only want to expand *some* of the 
%-constructions? In this case, %s for table and id, but not %W and 
such. Python apparently tries to convert them all and looks for 
matching variables in the tuple.

To get around this you can construct the SQL strings in multiple 
steps:

sql2 = ' from %s order by %s' % (table, id)
sql1 = "SELECT blah, DATE_FORMAT(date, \"%W, %d %b %Y\") " 

SQL = sql1 + sql2

# execute SQL...

This is only an example... not guaranteed to work as-is. But you'll 
get the idea.

HTH,

--Hans Nowak (zephyrfalcon at hvision.nl)
Homepage: http://fly.to/zephyrfalcon
You call me a masterless man. You are wrong. I am my own master.



From a_olme at my-deja.com  Fri Nov  5 02:05:18 1999
From: a_olme at my-deja.com (a_olme at my-deja.com)
Date: Fri, 05 Nov 1999 07:05:18 GMT
Subject: return boolean from functions
References: <7vrq9b$hmr$1@nnrp1.deja.com> <7vsvef$elo$1@nnrp1.deja.com>
Message-ID: <7vtvjd$4j9$1@nnrp1.deja.com>

My intention was to return true if a operation went well and false else.
But I wasn't sure that 0==false and <0> false. So thanks for the
answers.

//Anders Olme


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



From mgm at unpkhswm04.bscc.bls.com  Wed Nov  3 13:08:47 1999
From: mgm at unpkhswm04.bscc.bls.com (Mitchell Morris)
Date: 3 Nov 1999 18:08:47 GMT
Subject: exceptions from C
References: <Pine.LNX.4.10.9911040950440.25629-100000@crb.crb-web.com>
Message-ID: <slrn820uhe.n7i.mgm@unpkhswm04.bscc.bls.com>

Ordinarily I would sneer at your puny spam-trapped return address, but you
didn't ask for an email followup. Good for you!

In article <Pine.LNX.4.10.9911040950440.25629-100000 at crb.crb-web.com>,
Linux User wrote:
>I have been reading the section in the Lutz book on exceptions.  It seems
>that exceptions are no longer simply string objects.  Does this change the way
>exception are raised from C?  If so where might I find info on how to do this?
>
>Thanks,
>Wayne

I must admit I haven't look at my Lutz book in a while. Mea culpa.

You can find several examples of raising exceptions in the source distribution,
in the Demo directory. To raise one of the system exceptions, it's super-
easy:

	...
	if(some-test-failed) {
		PyErr_SetString(PyExc_KeyError, "bad dog! no biscuit!");
		return NULL;
	}
	...


To raise an exception already defined in a module, it's slightly more
complicated:

	PyObject* module = PyImport_AddModule("mymodule");
	PyObject* dict = PyModule_GetDict(module);
	PyObject* myexc = PyDict_GetItemString(dict, "MyException");
	// myexc is a borrowed reference ... don't decrement the count!
	PyErr_SetString(myexc, "bad dog! no biscuit!");



Finally, if you want to define the exception within the C++ part instead of
in the Python part:

	PyObject* module = PyImport_AddModule("mymodule");
    PyObject* dict = PyModule_GetDict(module);
    MYSTUFF_DataException = PyErr_NewException("MYSTUFF.DataException",
                                               NULL, NULL);
    PyDict_SetItemString(dict, "DataException, MYSTUFF_DataException);

and then use the code snippet above to raise it.


<exceptioning-for-fun-and-profit-ly y'rs>
+Mitchell


-- 
Mitchell Morris

My mind is like a steel trap - rusty and illegal in 37 states.
	-- Carrie Fish



From mhammond at skippinet.com.au  Mon Nov 15 17:15:04 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Mon, 15 Nov 1999 22:15:04 GMT
Subject: Python, COM and Excel
References: <80ps42$kkk$1@nnrp1.deja.com>
Message-ID: <IB%X3.4842$Xo.25297@news-server.bigpond.net.au>

charliehubbard76 at yahoo.com wrote in message <80ps42$kkk$1 at nnrp1.deja.com>...
>Hello,
>
>I'm trying to write a python script to take an excel
>workbook and turn it into html pages.  I'm trying to
>automate this process, but I can't seem to get the
>bounds of the selection through com.  I can get the
>contents of a selection with
>
>app = Dispatch(Excel.Application)
>app.Selection()
>
>But this just gives me the data contained with the cells.

Yes.  "app.Selection" is a range object.  Making a "call" on this object
returns the data in the range (ie, the "Value" property on the range object)

Try:
>>> app.Selection.Address
'$A$1:$C$5'

>I want to be able to check fonts, bold, and format
>information of each cell to get it as close to the excel
>formatting as possible.  In order to do this I need to

>>> app.Selection.Font.Name
'Arial'
>>>

etc.

Mark.





From viruswall at bam.com  Thu Nov 25 16:24:54 1999
From: viruswall at bam.com (viruswall at bam.com)
Date: Thu, 25 Nov 1999 16:24:54 -0500 (Eastern Standard Time)
Subject: InterScan NT Alert
Message-ID: <199911252130.QAA13063@python.org>

Receiver, Bell Atlantic Mobile has detected virus(es) in the e-mail attachment.

Date:  	Thu, 25 Nov 1999 16:24:54 -0500 (Eastern Standard Time)
Method:	Mail
From:  	<python-list at python.org>
To:    	<python-list at python.org>
File:  	Happy99.exe
Action:	clean failed - deleted
Virus: 	TROJ_SKA 



From ccgestjr at knoware.nl  Wed Nov 10 03:31:18 1999
From: ccgestjr at knoware.nl (cor gest jr)
Date: 10 Nov 1999 08:31:18 +0000
Subject: .dll and .exe file version informaton
References: <000101bf2b3c$a7d43e40$d32d153f@tim>
Message-ID: <87puxisq7d.fsf@cleopatra.clsnet.nl>

"Tim Peters" <tim_one at email.msn.com> writes:

> 
> [emuller at painewebber.com]
> > Can anyone give me an idea how to quikly get the version info on .dll
> > and .exe files under windows?
> 
> In Explorer, right-click on the file, select Properties from the context
> menu, and click the Version tab.
> 
> go-ahead-just-*try*-to-write-a-program-to-do-that<wink>-ly y'rs  - tim
> 
> 
> 
> 
ldd ?
A proper OS knows them, and can live with multiple versions.
nuf said..

cor
 
-- 
/* If GNU/LINUX has no solution, you have the wrong problem */ 
/* Q: $:which Linux     pa3eyp at amsat.org     A: $:ANY linux */
/*                     I prefer Slackware                   */ 
/*    http://utopia.knoware.nl/users/ccgestjr/linux.html    */



From spex66 at my-deja.com  Thu Nov 25 08:06:27 1999
From: spex66 at my-deja.com (spex66 at my-deja.com)
Date: Thu, 25 Nov 1999 13:06:27 GMT
Subject: decompress a bunch of .gz files
Message-ID: <81jc8i$ilp$1@nnrp1.deja.com>

Hy there,

I'm searching for a way to decompress all .gz
files in a directory
automatically.

GZIP and ZLIB documentation don't seem helpful to
me...

any hints?

thanks
Peter
(=PA=)



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



From sblakey at home.com  Mon Nov  8 15:05:08 1999
From: sblakey at home.com (Sean Blakey)
Date: Mon, 08 Nov 1999 20:05:08 GMT
Subject: Newbie question: Input
In-Reply-To: <XFMail.991108204005.mikael@isy.liu.se>
References: <XFMail.991108204005.mikael@isy.liu.se>
Message-ID: <Pine.LNX.4.20.9911081152400.26210-100000@seanb.sttln1.wa.home.com>

On Mon, 8 Nov 1999, Mikael Olofsson wrote:

> 
> On 08-Nov-99 Endgamer wrote:
>  > #####################################
>  >  Not a quote, just an on-the-spot observation:
>  >  You know you're a geek when your PC costs exactly
>  >  15 times as much as your stereo. And gives better
>  >  sound quality.<sigh>
>  >                                Endgamer
>  > #####################################
> 
> Sounds like we can define a geek index as 
> 
>        Price of PC
>   --------------------
>   15 * Price of stereo
> 
> as an analogue to the audiophile index
> 
>         Price of stereo
>   --------------------------
>   Price of record collection 
> 
> off-topic-again-ly y'rs
> 
> /Mikael
> 
> -----------------------------------------------------------------------
> E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
> WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
> Phone:   +46 - (0)13 - 28 1343
> Telefax: +46 - (0)13 - 28 1339
> Date:    08-Nov-99
> Time:    20:33:48
> 
> This message was sent by XF-Mail.
> -----------------------------------------------------------------------
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 
This formula does not take into acoount those of us who do not have
stereos (unless you count the speakers on my computer).  While I do have a
big ego, I do not believe that I deserve such a high geek index.  Perhaps
this would work better

def geekIndex(pc_cost, stereo_cost, default_index=1.0):
	try:
		return pc_cost / (15 * stereo_cost)
	except ZeroDivisionError:
		return default_index 


http://students.washington.edu/sblakey/
(206)934-5222
Public PGP key available at http://students.washington.edu/sblakey/public.asc






From bill.anderson at libc.org  Fri Nov 19 04:36:54 1999
From: bill.anderson at libc.org (Bill Anderson)
Date: Fri, 19 Nov 1999 02:36:54 -0700
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <14384.16479.836502.161209@anthem.cnri.reston.va.us> <80pifh$12n1$1@hub.org> <8s_X3.676$pw2.12189912@news.randori.com> <80r02c$1ho1$1@hub.org>
Message-ID: <38351A36.F41D2C58@libc.org>

"T. Muddletyn" wrote:
> 
> On Mon, 15 Nov 1999 12:55:58 -0800, Robin Dunn <robin at alldunn.com> wrote:
> >Could it be moved to the Starship or perhaps to a machine at CNRI to be part
> >of python.org?
> 
> Thanks to Gordon McMillan's initiative I have heard that Christian is willing
> to host Parnassus somewhere on the Starship. This may be a big help. However,
> if possible, i'd like to stay where I am, as I've have my little vex shell for
> quite a while and am comfortable and familiar with it. However due to various
> concerns about load/bandwidth raised here, I am discussing these things with
> the vex.net admin to see what he thinks of all this.

I could mirror it if it would help.
May or may not be immensely faster, but keeps the badnwidth usage down a
bit :-)
(... I presume it isn't massively huge ;-)


Bill
-- 
"They laughed at Columbus, they laughed at Fulton, they laughed at the
Wright brothers.  But they also laughed at Bozo the Clown." -- Carl
Sagan



From leonar_n at epita.fr  Mon Nov 15 09:55:00 1999
From: leonar_n at epita.fr (leon)
Date: Mon, 15 Nov 1999 15:55:00 +0100
Subject: PyApache
Message-ID: <80p6p4$or5$1@oceanite.cybercable.fr>

Hi,

I'm using PyApache, and I would like to know if it's possible to make
'persistent' modules. In fact I want to do only one "import file.py" like in
mod_perl, instead of loading this file each time the script is parsed.

Thanks.





From ivnowa at hvision.nl  Thu Nov  4 17:08:42 1999
From: ivnowa at hvision.nl (Hans Nowak)
Date: Thu, 4 Nov 1999 23:08:42 +0100
Subject: Module for generic platform information: platform.py
In-Reply-To: <38214ADD.1C7AB1CD@lemburg.com>
Message-ID: <199911042207.XAA16602@axil.hvision.nl>

On 4 Nov 99, M.-A. Lemburg wrote:

> There is a new release of the platform.py module available at:
> 
>         http://starship.skyport.net/~lemburg/platform.py
> 
> Still needed:
>     ? support for WinCE
>     ? support for MS-DOS (PythonDX ?)

Hmm... Doesn't that work already? I tried it several days ago, so 
with an older version of platform.py, obviously. Here's what it says 
when fed to Python-DX:

c:/djgpp/bin/uname.exe: invalid option -- p
Try `c:/djgpp/bin/uname.exe --help' for more information.
MS-DOS-7-pc

(OK, the error message doesn't belong there, but that's just because 
of the uname call. Most DOS users won't have an uname at all on their 
system. In spite of this, it does find the right DOS version. By the 
way, note that the os.popen, used to call uname, works. ;-)

When run with Python for Windows (on the same PC) it says:

Windows-95-4.00.1111

I don't know what it does on DOS 6 and lower, though. =/

--Hans Nowak (zephyrfalcon at hvision.nl)
Homepage: http://fly.to/zephyrfalcon
You call me a masterless man. You are wrong. I am my own master.



From fredrik at pythonware.com  Mon Nov 29 06:03:23 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 29 Nov 1999 12:03:23 +0100
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <81bmmi$279$1@nnrp1.deja.com> <3841EE77.6B5D6AC9@wjk.mv.com>
Message-ID: <01be01bf3a59$5ec0feb0$f29b12c2@secret.pythonware.com>

William J. King <wjk at wjk.mv.com> wrote:

(some quickies)

> 1.  like the indents -- makes coding easier to read and follow  -
>         requires discipline me thinks

well, I'm totally undisciplined, and I just love
indentation...

> 2. would like something like  'increment'/'decrement' operators
>         either in form:
>                     "x++"  or "incr x" vice "x = x + 1 " but its ok

if you find yourself using a lot of "x = x + 1" in your
code, you're probably not using for-in and other
sequence operations as much as you should...

> 3. would like to be able to iterate through a list using 2/more variables
> instead of
>         only one (unless its possible and I missed it) :
>                             for (x,y,z) in list:
>                                     pass            #favorite thing about
> Python

if list contains 3-tuples, that works exactly as you'd
expect.  if you have three lists, you can do:

    for x, y, z in map(None, list1, list2, list3):
        ...

a better syntax for this might make it into 1.6 (at least
according to GvR's presentation on the last python con-
ference).

http://www.foretec.com/python/workshops/1998-11/proceedings/guido/index.html

(hmm.  if I leave out the index.html part, the foretec
server redirects me to the fortec site ;-)

> 
> 4. conversion of integer to a string ( which may exist but havn't found
> yet) 

here are a few ways to do that:

    `integer`
    str(integer)
    "%d" % integer

> 5. would like to see regular array construct in addition to
> dictionary/list/tuples
>         faked that one with a dictionary

what's a "regular array construct", and how does that
differ from a python list?

(if you mean multidimensional arrays, numerical python
has that -- and those parts of numpy is very likely to
make it into standard python any day soon...)

> 6. some simpler description of current regular expressions in python
> some knowledge of grep/vi/awk/sed/perl/tcl - glob - regexp & regsub
> and now python can be a dangerous thing -- regular expressions rule if
> you get them right

yeah, but regular expression syntax suck ;-)

    Some people, when confronted with a problem, think
    "I know, I'll use regular expressions".  Now they have
    two problems.
        Jamie Zawinski, on comp.lang.emacs

(I hope Python will grow a much cooler string pattern
syntax some day, but that's another story.)

> 7. Tkinter is proving to be easier than I thought -- mostly intuitive from
> some previouse experience with Tcl/Tk - X-windows/Win95

with a little performance tuning, and some additional
widgets, Tkinter is quite ok (but I might be biased...)

...

</F>




From fredrik at pythonware.com  Tue Nov 30 03:32:26 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 30 Nov 1999 09:32:26 +0100
Subject: Htmlizing text
References: <943909426snz@vision25.demon.co.uk>
Message-ID: <020101bf3b0d$6fcc3210$f29b12c2@secret.pythonware.com>

Phil Hunt <philh at vision25.demon.co.uk> wrote:
> Is there a function in the standard Python library to HTML-ize text,
> i.e. to replace 'a > b & c < d' with 'a &gt; b &amp; c &lt; d'?

>>> import cgi
>>> cgi.escape('a > b & c < d')
'a &gt; b &amp; c &lt; d'

> (I realize this can be done with regular expressions, but is there
> an easy way?)

hmm.  doing this with regular expressions seems a bit
backwards.  but I might be missing something...

(for info on how to go from escaped entities to plain
text, see the 'htmlentitydefs-example-2.py' script in
the eff-bot guide).

</F>

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




From mhammond at skippinet.com.au  Sun Nov  7 00:23:58 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sun, 07 Nov 1999 05:23:58 GMT
Subject: Doing it with Python, Windows SEND-TO and *.cpl-files
References: <8024lh$2mnb$1@bnews.gigabell.net>
Message-ID: <O18V3.5031$pf4.18246@news-server.bigpond.net.au>

Florian W. wrote in message <8024lh$2mnb$1 at bnews.gigabell.net>...
>When I push a script-into the Send-To-Folder,
>what must be done, to get the filename which was selected with it ?

Should be in sys.argv[1]

>Is it possible to create a *.cpl-file ?
Not at things currently stand.  However, a .cpl file is just a .dll with
some special entry points, so it would not be hard to make a Python
interface to this.

>What is the easiest way to manage dialogs ? (With dialog-boxes and all
>rc-file-stuff)

There are lots of GUI toolkits available.  If you have a .rc file, the best
solution for them is probably Pythonwin (however, if you want dynamic
dialogs defined inside your code, Pythonwin is not the simplest)

Mark.





From prestonlanders at my-deja.com  Fri Nov  5 09:16:03 1999
From: prestonlanders at my-deja.com (Preston Landers)
Date: Fri, 05 Nov 1999 14:16:03 GMT
Subject: return boolean from functions
References: <7vrq9b$hmr$1@nnrp1.deja.com> <7vsvef$elo$1@nnrp1.deja.com>
Message-ID: <7vuoqt$m79$1@nnrp1.deja.com>

In article <7vsvef$elo$1 at nnrp1.deja.com>,
  Preston Landers <prestonlanders at my-deja.com> wrote:

> Alternatively, you can do this:
>
> TRUE=1
> FALSE=0
>
> def foo(bar):
>   if bar == "spam":
>      return TRUE
>   else:
>      return FALSE

Sorry to follow up my own article, but it should be noted that the way I
have it above will cause a namespace lookup (TRUE and FALSE) every time
you call foo() (aside from the usual function overhead.)  So, if you are
calling this function from a critical loop, you might want to avoid the
namespace lookup and just return None or 1.  Also note that the word
"None" is optional when you want to return false; you can do this:

  if foo:
     return  # will return None

anders, I noticed your other message about returning true for success
and false for fail.  That's fine if that works for you, but FYI the
convention in Unix[1] is generally to return false for success, and true
for an error condition.  This way the exact condition can be contained
in the return value, such as:

def foo():
   return "foo is bad"

The code that calls this will look like this:

result = foo()
if result:
   print "Foo failed and said: %s" % result
   raise SystemExit(1)

# continue on merry way

---Preston

[1] Assuming you think Unix is conventional


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



From aahz at netcom.com  Thu Nov 11 16:44:44 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 11 Nov 1999 21:44:44 GMT
Subject: What are Aspects? (Re: How can I make my assertions smarter?)
References: <Pine.WNT.4.05.9911032042440.109-100000@david.ski.org> <slrn82h5o6.26j.wtanksle@hawking.armored.net> <80eocf$8rc$1@nntp5.atl.mindspring.net> <WIDW3.331$Ae3.219038@news-dal.corridex.com>
Message-ID: <80fdcc$d4f$1@nntp6.atl.mindspring.net>

In article <WIDW3.331$Ae3.219038 at news-dal.corridex.com>,
Evan Simpson <evan at tokenexchange.com> wrote:
>
><disclaimer>All I know about Aspects I learning in a half-hour of
>browsing</disclaimer>

Okay, but I'm still going to ask questions.

>I think that this depends on the language in which you write your Aspects (I
>don't mean C/Python/etc, I mean the special Aspect-specification language).
>In the examples I looked at, pattern-matching on the class, method, and
>method arguments determined where each part of an Aspect would be applied.

So it's essentially a kind of pre-processor?

(I'm reasonably bright in my own way, but I seem to only be able to
learn new concepts by relating them to ones I already know -- I can't
just accept something on its own merits.)
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From aaron_watters at my-deja.com  Wed Nov 10 22:56:56 1999
From: aaron_watters at my-deja.com (aaron_watters at my-deja.com)
Date: Thu, 11 Nov 1999 03:56:56 GMT
Subject: Just a note
References: <381D5E84.98931682@bibsyst.no> <381D78C3.9BBECCB0@bibsyst.no> <7vut5l$pgn$1@nnrp1.deja.com> <3826A8E2.D2C557D7@bibsyst.no> <LhzV3.9524$Xd2.519654@typhoon.nyroc.rr.com>
Message-ID: <80deq8$10f$1@nnrp1.deja.com>

In article <LhzV3.9524$Xd2.519654 at typhoon.nyroc.rr.com>,
  "Darrell" <news at dorb.com> wrote:
> I believe Aaron was just throwing out ideas since you requirements are
> vague. In most cases C++ solution will be faster than pure Python.
Sounds
> like you want a disk based dictionary. I'd be curious how much C++
will buy
> you in this case. Disk I/O will eat into your gains.

Actually I wasn't just throwing out ideas.  In the case of bplustree.py
the overhead of disk IO should dominate the computation, so it might
be a pretty good solution.  Could be wrong tho.

bplustree is a classical b+tree implementation, more or less, and
it may be within a small constant factor of a C++ implementation, or
it might not.  In theory it could be.  Please try it out and correct
me.
   -- Aaron Watters http://www.chordate.com

===
A PhD in Theory, is a PhD, in theory.  -- folklore


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



From ajung at sz-sb.de  Sat Nov  6 05:35:37 1999
From: ajung at sz-sb.de (Andreas Jung)
Date: Sat, 6 Nov 1999 11:35:37 +0100
Subject: pybench...where ?
Message-ID: <19991106113537.A17850@sz-sb.de>

I remember that there has been a PyBench suite available
some time ago. However I can not find the current version.

Any hint ?

Thanks,
Andreas



From tim_one at email.msn.com  Wed Nov  3 00:08:56 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 3 Nov 1999 00:08:56 -0500
Subject: __str__ vs. __repr__
In-Reply-To: <5lln8ht38a.fsf@eric.cnri.reston.va.us>
Message-ID: <000401bf25b9$88ea57e0$2e2d153f@tim>

[Randall Hopper wonders about str-vs-repr, Tim explains that lists ask
 their elements to produce repr() even if the list itself was passed
 to str()
]

[Guido]
> Actually, Python's internals *do* follow the convention.  Dicts and
> lists don't define __str__(), so str() for them defaults to
> __repr__(), and then of course repr() is used for the items.  It may
> be confusing, it may not be what you want, but it *is* consistent ;-(

Not to mention undocumented <wink>.

[Randall]
> For consistency, would it make sense to change this for Python 1.5.3 (that
> is, have sequence and dict types pass 'str-vs-repr'ness down)?

[Guido]
> This has been asked a few times before (and usually it's been reported
> as a bug, which it isn't -- see above).  I happened to see this post
> and it made me delve deep into my intuition trying to figure out why I
> don't like propagating str() down container items.
>
> Here's what I think it is.  There's no reason why an object's str()
> should be particularly suited to being included in list syntax.

This seems much more a consequence of the current design than an argument in
favor of it.  That is, had Python been designed so that the builtin
container types did "pass down" str-vs-repr'ness, an object's str() would
have every reason to produce a string suited to being etc.

> For example, I could have a list containing the following items:
>
> 	1		# an integer
> 	'1'		# a string (of an integer literal)
> 	'2, 3'		# a string containing a comma and a space
> 	'], ['		# a string containing list delimiters
>
> Under the proposed rules, this list would print as:
>
> 	[1, 1, 2, 3], []
>
> I would find this confusing

Me too, but I find the current design more rigidly consistent than useful
(see below).  In a world where containers passed str() down, a container's
str() would presumably be responsible for adding disambiguating delimeters
to element str() results when needed (the container knows its own output
syntax, and can examine the strings produced by its elements -- not rigidly
consistent, but useful <wink>).

> and I worry that it could be used to fool the user.

People can already define __repr__ to return anything whatsoever; the
reports of people getting fooled by this are conspicuous by absence <wink>.

Here's something wholly typical of what I dislike:

>>> from Rational import Rat, Format
>>> Rat.set_df(Format.Format(mode=Format.FIXED, prec=3, use_tag=0))
Format(mode=Format.MIXED, prec=8, base=10, round=Round.NEAREST_EVEN,
       use_tag=1, use_letters=1)
>>> one_tenth = Rat.Rat(.1)
>>> one_tenth
Rat(3602879701896397L, 36028797018963968L)
>>> print one_tenth
0.100
>>>

That is, in interactive mode, I'm forever using "print" because the default
of applying repr() to raw expressions produces the output least useful in
interactive hacking (I don't care about reproducing the object exactly from
the string when typing a raw expression at the prompt!  The mental ratio of
two giant integers isn't helpful here.).

Carry it one more step, and nothing simple suffices anymore:

>>> values = [one_tenth, one_tenth + 100]
>>> values
[Rat(3602879701896397L, 36028797018963968L),
 Rat(3606482581598293197L, 36028797018963968L)]
>>> print values
[Rat(3602879701896397L, 36028797018963968L),
 Rat(3606482581598293197L, 36028797018963968L)]
>>>

So I'm forever typing this instead:

>>> map(str, values)
['0.100', '100.100']
>>>

Throw a dict into it, and it's hopeless:

>>> recip = {one_tenth: 1/one_tenth, 1/one_tenth: one_tenth}
>>> print recip
{Rat(3602879701896397L, 36028797018963968L):
     Rat(36028797018963968L, 3602879701896397L),
 Rat(36028797018963968L, 3602879701896397L):
     Rat(3602879701896397L, 36028797018963968L)}
>>>

Having gone thru the same business in dozens of classes over the years, I
find the current design simply unusable in interactive mode.  For a while I
defined just __str__, bound __repr__ to that too, and added a .repr()
*method* for the unusual cases in which I really needed a faithful string.
But that frustrated other code that expected explict repr() calls, and/or
the `` notation, to produce the long-winded version.  So that sucked too.

It's even an irritation sticking to builtin types; e.g., here assuming
Latin-1 comes across intact:

>>> names = ["Fran?ois", "Tim"]
>>> print names[0]
Fran?ois
>>> >>> print names
['Fran\347ois', 'Tim']
>>>

That isn't helpful either -- it's frustrating.

in-the-face-of-ambiguity-refuse-the-temptation-to-do-the-opposite-of-
    what-the-user-asked-for<0.9-wink>-ly y'rs  - tim





From donn at u.washington.edu  Fri Nov  5 18:00:23 1999
From: donn at u.washington.edu (Donn Cave)
Date: 5 Nov 1999 23:00:23 GMT
Subject: __str__ vs. __repr__
References: <000001bf275c$02370d20$1d2d153f@tim>
Message-ID: <7vvni7$19q0$1@nntp6.u.washington.edu>

Quoth "Tim Peters" <tim_one at email.msn.com>:

| Let me back off to what repr and str "should do":
|
| repr(obj) should return a string such that
|     eval(repr(obj)) == obj
| I'd go further and formalize a property that's currently honored but
| implicitly:  the string returned by repr should contain only characters c
| such that
|     c in "\t\n" or 32 <= ord(c) < 127
| i.e. it should be restricted to the set of 7-bit ASCII characters that the C
| std guarantees can be read back in faithfully in text mode.  Together, those
| properties spell out what repr() is "good for":  a highly portable and
| eval'able text representation that captures *everything* important about an
| object's value.

For real?  I'm very used to seeing repr's that don't eval at all -

'<socket object, fd=3, family=2, type=1, protocol=0>'
"<open file '<stdout>', mode 'w' at 80005660>"
'<exceptions.NameError instance at 80048ae0>'

Now it's fine with me if the author does choose to implement repr that
way, it's just hard to imagine it could be meaningfully implemented by
all objects, or that there's any purpose in trying.

Well, really I guess I would prefer that authors _not_ try to capture
*everything* important, if that means a large data dump.  repr is used
in lots of places, like the interactive interpreter, where the output
only has to be roughly indicative, and I have never seen and can't find
any example of its use with eval or in any context where it has to be
definitive.

If there are folks who want to provide this functionality at the expense
of making a string that's unusable for repr's normal applications, 
they should put it in a different function.  

| In contrast, str() should return a string that's pleasant for people to
| read.  This may involve suppressing obscure or minor details, displaying it
| in a form that's inefficient-- or impossible --to process, or even plain
| lying.  Examples include the Rat class I used before (the ratio of two huge
| integers is exact but unhelpful), a DateTime class that stores times as
| seconds from the epoch (ditto), or a Matrix class whose instances may
| contain megabytes of information (and so a faithful repr() string may be
| intolerable almost all the time).

Bah.  If we get str() because it's pleasant to read, aren't we worried
that repr() will be harsh and stressful?  Will one repr() too many send
one of us into a homicidal rage one of these days?

Let us think about taming the cruel repr()!  If I may speculate on what
you're trying to say, str() is pleasant because it shows the data in
form that has been digested for human consumption.  In contrast to
repr(), which is intended for the machine.

But as I don't agree with with the latter, I don't have to agree with
the former either.  On the contrary, it's str() that's intended for the
machine, and repr() for the human.  Consider rfc822, one of the three
Python modules in the standard distribution that actually implements
__str__:

>>> fp = open('mimetools', 'r')
>>> m = rfc822.Message(fp)
>>> str(m)
'Path: news.u.washington.edu!logbridge.uoregon.edu!news.mind.net!not-for
-mail\012From: dragondm at integral.org (The Dragon De Monsyne)\012Newsgrou
ps: comp.lang.python\012Subject: RFC complient mimetools...\012Date: 29 
Sep 1999 13:27:03 GMT\012Organization: InfoStructure - Ashland, OR\012Li
nes: 443\012Sender: dragondm at polar.integral.org\012Message-ID: <7st437$r
9s$2 at utah.laird.net>\012NNTP-Posting-Host: ip202.mind.net\012Mime-Versio
n: 1.0\012Content-Type: text/plain; charset=us-ascii\012X-Newsreader: kn
ews 1.0b.0\012Xref: news.u.washington.edu comp.lang.python:68258\012'

Pleasant, eh? [wink, wink, nudge, nudge]  I wrapped the lines, in case
the original line length of 529 might hurt someone's software.

So look at the way it's really used.  repr is for us, str is for the
machine.  Now for sure, there are quite a number of objects out there
out there that don't repr as readable as they could, but I think very
few of them are that way because they didn't get a chance to use their
str instead.

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu



From b2blink at hotmail.com  Fri Nov 26 09:00:17 1999
From: b2blink at hotmail.com (Ulf Engstrøm)
Date: Fri, 26 Nov 1999 09:00:17 CET
Subject: SMTP - again ;)
Message-ID: <19991126080017.26482.qmail@hotmail.com>

I've gotten my prog to work now, only have some problems with our 
server/firewall, but that's hardly an issue.

Anyway, something else came up, when I enter a name in the header:
"To: Ulf Engstr?m" the mail gets to the dest as:

To: Ulf.Engstr?m at mb04.swip.net if I use mb04.swip.net as outgoing server, 
eventhough the address is completely different...and the add before @ is not 
correct either, since it's the Real Name..
Is there a way to get rid of that? In the To-field of the mail I just want 
the name, not @anywhere...

Thanx for all previous help :)
Best regards

Ulf Engstr?m

http://www.b2b-link.com
ulf.engstrom at b2b-link.com

______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com



From malcolmt at smart.net.au  Tue Nov  2 03:25:15 1999
From: malcolmt at smart.net.au (Malcolm Tredinnick)
Date: Tue, 2 Nov 1999 19:25:15 +1100
Subject: commas vs. spaces. vs. the legions of new lines
In-Reply-To: <99110201025204.01264@quadra.teleo.net>; from Patrick Phalen on Tue, Nov 02, 1999 at 12:59:43AM -0800
References: <1270647588-25358408@hypernet.com> <007f01bf250c$72dd3c40$f29b12c2@secret.pythonware.com> <99110201025204.01264@quadra.teleo.net>
Message-ID: <19991102192515.A1698@Ridcully.home>

On Tue, Nov 02, 1999 at 12:59:43AM -0800, Patrick Phalen wrote:
> This is quite possibly the scariest idea I have ever heard.

Just a quick post before the Off-Topic Police arrest me. I think that in
general this is a good idea (particularly the bit about having fun). Think
of how many time people try to do things the Unix way on here and then find
out that that option isn't implement in the Windows version. Having a
reasonably complete suite of the common tools available in perl (or
Python, just to keep some relevance) saves a lot of hassle.

Cheers,
Malcolm Tredinnick




From aj10 at my-deja.com  Tue Nov 30 13:20:40 1999
From: aj10 at my-deja.com (aj10 at my-deja.com)
Date: Tue, 30 Nov 1999 18:20:40 GMT
Subject: Recursion?
References: <38421570.9C3D9AEC@videotron.ca>
Message-ID: <8214hi$q87$1@nnrp1.deja.com>

In article <38421570.9C3D9AEC at videotron.ca>,
  Pierre-Yves Cloutier <pycloutier at videotron.ca> wrote:
> Can I use recursion in python?
>
> thanks
>
> Pierre-Yves Cloutier
>
>
An example:
--
Python 1.5.2 (#0, Apr 13 1999, 10:51:12) [MSC 32 bit (Intel)] on win32
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> def f(i):
...     print i
...     if i==0: return
...     f(i-1)
...
>>> f(4)
4
3
2
1
0
>>>
--


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



From a.eyre at optichrome.com  Tue Nov 16 12:05:44 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Tue, 16 Nov 1999 17:05:44 -0000
Subject: function override???
In-Reply-To: <80s0sd$1g4k$1@rtpnews.raleigh.ibm.com>
Message-ID: <004501bf3054$d1828040$3acbd9c2@peridot.optichrome.com>

> ;---fprot.py
> def overrideme(self, str, someint)
>
> ;--fimp.py
> import fprot   #I know this ain't right.
>                           # I don't know how to import from other python
> scripts yet.  Enlighten me?
>
> def overrideme(self, str, someint)
>     do_some_junk()

Probably easier to use a couple of classes:

###
# interface.py
class Interface:
	def callback1(self, args):
		pass

###
# implementation.py
from interface import Interface

class Implementation(Interface):
	def callback1(self, args):
		do_some_junk()

###

Then in the C code do something like:

// N.B. This will leak a bit. Need to DECREF some stuff...
PyObject* mod = PyImport_ImportModule("implementation");
PyObject* dict = PyModule_GetDict(mod);
PyObject* imp_class = PyDict_GetItemString(dict, "Implementation");
PyObject* imp_inst = PyEval_CallObject(imp_class, PyTuple_New());
PyObject* callback1 = PyObject_GetAttrString(imp_inst, "callback1");
PyObject* result = PyEval_CallObject(callback1, Py_BuildValue("(s)",
"arg"));

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com
--------------------------------------------




From cfelling at iae.nl  Thu Nov 25 18:09:32 1999
From: cfelling at iae.nl (Carel Fellinger)
Date: 26 Nov 1999 00:09:32 +0100
Subject: Newbie: raw_input and other input methods
References: <19991125.1481400@adam-aurag.penguinpowered.com>
Message-ID: <81kfjc$1vg$1@vvs.superst.iae.nl>

Hassan Aurag <aurag at geocities.com> wrote:

> Anyway, I used raw_input to be able to use readline and its nifty back 
> arrow, up arrow...

> However when I wanted to have a nifty prompt like the following one:

> raw_input("\033[0;34mInput: \033[0m"), I was surprised it got all 
> screwed up.

bash copes with this, so I dived into the source. And low and behold
there it is: bash uses the readline special codes RL_PROMPT_START_IGNORE
(='\001') and RL_PROMPT_END_IGNORE (='\002') to tell readline to 'ignore'
the chars in between.  Using this info you can get what you want:)

e.g.
>>>raw_input('\001\033[0:34m\002Input: \001\033[0m\002')

works like a charm, atleast here on my linux box




-- 
groetjes, carel



From irwin at mail.com  Wed Nov 10 05:55:19 1999
From: irwin at mail.com (Roger Irwin)
Date: Wed, 10 Nov 1999 11:55:19 +0100
Subject: ASCII delimited files
Message-ID: <942231502.1914033961@news.interbusiness.it>

Is there any function or module available for parsing ASCII delimited files,
before I go and re-invent the wheel writing my own.

BTW, I know it is simple to do, but I was thinking along the lines that a
'standard' could easily be substituted for a C version for large files.




From gerrit.holl at pobox.com  Wed Nov 10 14:40:23 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Wed, 10 Nov 1999 20:40:23 +0100
Subject: Intelligent strptime?
In-Reply-To: <14377.41778.712189.7874@gargle.gargle.HOWL>
References: <19991110171256.A5196@optiplex.palga.uucp> <14377.41778.712189.7874@gargle.gargle.HOWL>
Message-ID: <19991110204023.A8220@optiplex.palga.uucp>

Charles G Waldman wrote:
> Gerrit Holl writes:
>  > 
>  > Does there exist an intelligent strptime? "26/11/99", for example, can _only_
>  > mean the 26th November of this year, and the chance that "14:11:12" means
>  > 14 november 1912 is very small. Is there a module/function that guesses the
>  > format? There isn't doubt very often, only for 12/11/99, US is 11 december,
>  > Europe is 12 November (isn't it?).
> 
> I think that there's an ambuguity between US-style MM/DD and
> European-style DD/MM whenever the day of the month is 12 or less.   Of 
> course when DD==MM then the ambuguity doesn't matter (e.g. 3/3/99).
> But that still leaves 11 days in each month which have this ambiguity, 
> which comes to 132 days per year.  So, given a 365-day year, 36% of
> the dates are ambiguous, which doesn't make "guessing" look like a
> very attractive option to me.

It could return a tuple or something then. But if I want people to choose:
either input "Nov, 23" or "12/6/99" or "1 December 2000", I don't want to
check too much. There are much date processing programs (date, I think) that
do this intelligent. You could also let the above think depend on the timezone,
if it's an american timezone, take the american way, else, take the
european way. But returning a tuple on 132 days seems better to me.

regards,
Gerrit.
-- 
linuxgames.nl.linux.org	 All about games on Linux in Dutch (under construction).
www.nl.linux.org	       The Dutch resource for Dutch Linux documentation.
www.nl.linux.org/discoverb		    Learn foreign words and definitions.
www.nl.linux.org/~gerrit/asperger	 I'm an Asperger, go here for more info.
www.nl.linux.org/~gerrit			    About me and pictures of me.



From fdrake at acm.org  Fri Nov 12 10:08:49 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Fri, 12 Nov 1999 10:08:49 -0500 (EST)
Subject: Tkinter and after
In-Reply-To: <00a001bf2d1f$fcd3ac70$f29b12c2@secret.pythonware.com>
References: <382C2854.E9494EFE@lri.fr>
	<14380.11058.151680.518510@weyr.cnri.reston.va.us>
	<00a001bf2d1f$fcd3ac70$f29b12c2@secret.pythonware.com>
Message-ID: <14380.11649.578901.273647@weyr.cnri.reston.va.us>

Fredrik Lundh writes:
 > what version are you using?  if you
 > run this script, can you move the
 > scale widget after you've clicked

  Mea Culpa!  I was sure we'd been using after() in Grail; I guess not 
where I'd thought we used it.
  I guess I'm a little too fast on the keyboard this morning!
  Everyone please ignore my false post!  Sorry 'bout that.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From ejohnson at netcom.com  Tue Nov 16 23:45:55 1999
From: ejohnson at netcom.com (Eric Johnson)
Date: 17 Nov 1999 04:45:55 GMT
Subject: Bug? - win32pipe.popen2 is moody under W98
References: <80qkbv$2m5$1@nntp5.atl.mindspring.net> <cg6Y3.5073$Xo.26549@news-server.bigpond.net.au>
Message-ID: <80tbu3$923$1@nntp1.atl.mindspring.net>

"Mark Hammond" <mhammond at skippinet.com.au> writes:

>Eric Johnson wrote in message <80qkbv$2m5$1 at nntp5.atl.mindspring.net>...

>>Can anyone say either way whether this can be expected to work under
>>W98?

>Unfortunately it is not.  This is mentioned in the Python FAQ (including a
>link to the MS KB article that documents the bug)

>On the brighter side, Bill Tutt finally decided to work around this bug in
>Windows 98.  It will be there in the next version of win32all.  In the
>mean-time, I could probably send you the necessary updates... 

An upcoming work around in win32all is sufficient for me.  I'll keep
an eye out for it on your site and happily abuse it when it appears.

-Eric



From robin at jessikat.demon.co.uk  Tue Nov  2 06:41:33 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Tue, 2 Nov 1999 11:41:33 +0000
Subject: Tcl82.dll not found
References: <MbKS3.4152$IT5.272573@ndnws01.ne.mediaone.net>
 <7vg3hd$75q$1@newton.pacific.net.au>
 <NwWS3.4286$IT5.283485@ndnws01.ne.mediaone.net>
Message-ID: <1J7QPPAt3sH4EwHa@jessikat.demon.co.uk>

In article <NwWS3.4286$IT5.283485 at ndnws01.ne.mediaone.net>, Art
Goldhammer <agold1 at mediaone.net> writes
>
>
>----- Original Message -----
>From: Kiyo Kelvin Lee <kiyolee*remove*@ctimail.com>
>
>> You must use Tcl8.0.5 included in Python 1.5.2. It's somehow hard-linked
>to
>> use Tcl80.dll/Tk80.dll. Tcl8.0.5 and Tcl8.2 can coexist in a system. If
>you
>> really want to use Tcl82.dll/Tk82.dll, you have to recompile Python
>> yourself.
>
>Thanks. I tried installing Tcl8.0.5 in its own directory and putting it in
>the path. Tkinter now works, and IDLE runs, which it didn't do before. But
>when I exit IDLE I get a page fault in kernel32.dll from which there is no
>escape but a hard reboot. Is this a Tcl conflict or a known problem with
>IDLE? -- Art
>
>
>
>
I have managed to get tkinter working with 1.5.2 with help from others
on clp. Perhaps there should be a repository for various pyd's like
this.
-- 
Robin Becker



From akuchlin at mems-exchange.org  Tue Nov 30 15:09:18 1999
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: 30 Nov 1999 15:09:18 -0500
Subject: python test frameworks
References: <slrn83u3um.4ge.quinn@groat.ugcs.caltech.edu> <14402.65296.413573.98151@weyr.cnri.reston.va.us>
Message-ID: <3d903fhhch.fsf@amarok.cnri.reston.va.us>

"Fred L. Drake, Jr." <fdrake at acm.org> writes:
>   There's also the somewhat-general (and completely undocumented)
> "test" package in the Python standard library.  That might be worth a
> look; I'm not sure how re-targetable it is.

It's usable for external software in 1.5.2, since I use it for the
test suite in the standalone distribution of the PCRE module.  

--amk



From bernhard at alpha1.csd.uwm.edu  Mon Nov 15 19:14:04 1999
From: bernhard at alpha1.csd.uwm.edu (Bernhard Reiter)
Date: 16 Nov 1999 00:14:04 GMT
Subject: Line Benchmarks Tkinter vs. wxPython
References: <01bf2fa7$02992ca0$8a42c3d1@Cyberus.cyberus.ca>
Message-ID: <slrn8318ec.9dg.bernhard@alpha1.csd.uwm.edu>

On Mon, 15 Nov 1999 20:22:16 GMT, Gordon Williams <g_will at cyberus.ca> wrote:
>I have done some benchmark testing to see the difference in speed between
>plotting multi point lines using Tkinter and wxPython.  
Very nice. I appreciate the effort.

I couldn't figure out on which window platform you tried this.

wxGtk or wxWindows?

as the wxWindows compatible libraries use native windows calls
this might make a significant difference.

	Bernhard
-- 
Research Assistant, Geog Dept UM-Milwaukee, USA.  (www.uwm.edu/~bernhard)
Free Software Projects and Consulting 		         (intevation.net)  
Association for a Free Informational Infrastructure            (ffii.org)



From aa8vb at yahoo.com  Wed Nov 24 09:57:53 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Wed, 24 Nov 1999 09:57:53 -0500
Subject: Absolute newbie question
In-Reply-To: <Pine.SOL.3.96.991124165729.11616B-100000@giasbma>
References: <Pine.SOL.3.96.991124165729.11616B-100000@giasbma>
Message-ID: <19991124095753.A1432177@vislab.epa.gov>

Philip Tellis:
 |Where are the man pages for python?  I am new to python, and would like to
 |look through the man pages before I start.  Someone please help.

Fredrik pointed you to the man page, but what you may want is the
full library and language documentation.  Bookmark this URL:

    http://www.python.org/doc/

Tutorial is where to start.  Library Reference has all the goodies you can
import and use.

Randall



From lm123 at softhome.net  Wed Nov 24 00:17:01 1999
From: lm123 at softhome.net (lm123)
Date: Wed, 24 Nov 1999 10:47:01 +0530
Subject: (no subject)
Message-ID: <001801bf363b$3494ca80$539909ca@lalitm>

subscribe
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/19991124/34d83a4c/attachment.html>

From doliver at bc.sympatico.ca  Tue Nov 23 20:49:42 1999
From: doliver at bc.sympatico.ca (Don Oliver)
Date: Tue, 23 Nov 1999 17:49:42 -0800
Subject: Upgrade to version 1.5.2 - won't configure?
References: <Ql2_3.63$Lf2.352159@news.bctel.net> <383990AC.A7D2BD27@python.net> <EEl_3.52$jf7.366030@news.bctel.net> <383A82DE.29ED626F@python.net>
Message-ID: <KzH_3.64$Ge5.397561@news.bctel.net>

Thanks again to Tom and Michael for their help, as well as some good advice.
I don't intend to be writing C programs; only Python.
I'm probably trying to do too many new things at once, and should just use
the rpm's and defaults until I'm more experienced.
The Python web site doesn't say that all these extra programs are required,
nor does the readme in the download.



Thomas A. Bryan wrote:
> A reasonable goal...one that I don't attempt to acheive on my home box.
> (Disk space is fairly cheap, and I'd prefer to be developing than setting
> up a "nice" development machine.)

Sounds sensible to me. :-)

> Wow!  I never have any idea what's going on with Windows.  As far as I can
> tell, I run the install program and several megabytes of DLLs are dumped
> somewhere under the Windows subdirectory.  I seldom seem to have any
choices
> about what goes where, and there's no way to figure out, as far as I know,
> what files were installed by which programs.  RPM has powerful query
> capabilities, and I love it for that.

I said that I knew what I was doing, I never claimed to know all about the
DLL's, etc.
By that, I meant that I had no problem installing and using any programs.
Big difference!

***
Michael Hudson wrote:

>Don't we all... seriously, one thing I have learnt over my year or two
>with Linux is to let rpm do it's stuff and not move stuff around
>underneath it. It's possible to do it yourself using tarballs and mv
>and cp, but it becomes harder to keep track of what's where.

I guess I'll stick with rpms for now.

Thanks again to both.

Don Oliver
dayo_nanaimo at hotmail.com






From maxm at normik.dk  Wed Nov 24 05:08:45 1999
From: maxm at normik.dk (maxm)
Date: Wed, 24 Nov 1999 11:08:45 +0100
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <81g8po$reo$1@mach.vub.ac.be>
Message-ID: <002901bf3663$e4b62df0$11e9efc2@normik.dk>

> Mark Carroll (markc at chiark.greenend.org.uk) wrote:
> [snip]
> : Basically, I'd be really interested to learn: what _don't_ people like
> : about Python?
> [snip]

I miss deeply the ability to write:

i += 3
Text += "Some new text\n"

etc.

and the 

BigUn = (x>y) ? x : y

Those would be really nice syntax sugaring.

Gimme some sugar please!!!

:-)





From a.eyre at optichrome.com  Fri Nov  5 11:32:21 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Fri, 5 Nov 1999 16:32:21 -0000
Subject: Drawing (directed) graphs
In-Reply-To: <19991105160912.502.rocketmail@web602.mail.yahoo.com>
Message-ID: <000e01bf27ab$5548aba0$3acbd9c2@peridot.optichrome.com>

>> How can there be. Some OSs (Minix) don't support
>> graphics at all...

> I didn't say EVERY OS, just a growing number of the
> common ones (and of course the GUI libraries too).  

Sorry. Pedant mode cancelled.

However, AFAIK, every platform which Python runs on
supports ASCII text output, which is why I initially
suggested it as a method for drawing graphs.

> Tkinter is widely supposed to be platform-independent
> but that won't run on Minix either...

Pedant mode engaged. ;-)

I think that's called 'multi-platform' not 'platform-
independent'.

Platform independent code would be something like:

#!/usr/bin/env python

a = 1 + 1

Not terribly useful, but platform independent (at least
as far as Python is concerned.)

As soon as you import Tkinter, you lose a few platforms.
If you import posix or msvcrt, you lose more. 

This is why it annoys me that some of the scripts in the
standard Python library are not platform independent, but
rely on non-portable features of posix.

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From fredrik at pythonware.com  Wed Nov 17 08:11:04 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 17 Nov 1999 14:11:04 +0100
Subject: Dictionary???
References: <80u8se$1i1c$1@rtpnews.raleigh.ibm.com>
Message-ID: <009a01bf30fd$65b36ce0$f29b12c2@secret.pythonware.com>

Arinte <shouldbe at message.com> wrote:
> I looked in the tutorial and they gave a general overview.  It show how to
> view and delete items in a dictionary, but how do you add a pair to the
> dictionary?

http://www.python.org/doc/current/tut/node7.html

section 5.4:

    Here is a small example using a dictionary: 

    >>> tel = {'jack': 4098, 'sape': 4139}
    >>> tel['guido'] = 4127
    >>> tel
    {'sape': 4139, 'guido': 4127, 'jack': 4098}
    >>> tel['jack']
    4098
    >>> del tel['sape']
    >>> tel['irv'] = 4127
    >>> tel
    {'guido': 4127, 'irv': 4127, 'jack': 4098}
    >>> tel.keys()
    ['guido', 'irv', 'jack']
    >>> tel.has_key('guido')
    1

also see:
http://www.python.org/doc/current/lib/typesmapping.html

</F>




From jbauer at rubic.com  Thu Nov 18 04:28:34 1999
From: jbauer at rubic.com (Jeff Bauer)
Date: Thu, 18 Nov 1999 03:28:34 -0600
Subject: Python in Windows CE?
References: <19991118075134.9052.qmail@hotmail.com>
Message-ID: <3833C6C2.74797F41@rubic.com>

>> Has anybody used Python for development under Windows CE?

> Well, not really using it yet, but I've been trying to 
> gather some info about it, cause by next week I'll start
> doing some Python on a small machine running CE. I'd also
> be interested in all info that anyone might have and
> I'm glad to answer any questions about my development
> once I get it going.

Ulf, Rick, etc.

You will probably want to check out the Python CE
mailing list:

   http://www.egroups.com/group/python-ce/

Python works great on the CE platform. The chief
shortcoming is the lack of a friendly GUI, but
this issue is being addressed.  Check the mailing
list for further details.

-Jeff Bauer



From a.eyre at optichrome.com  Fri Nov 26 09:43:08 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Fri, 26 Nov 1999 14:43:08 -0000
Subject: SMTP - again ;)
In-Reply-To: <19991126080017.26482.qmail@hotmail.com>
Message-ID: <000301bf381c$8e6337d0$3acbd9c2@peridot.optichrome.com>

> "To: Ulf Engstr?m" the mail gets to the dest as:

This is not a valid RFC822 value. Read RFC822. It will use...

> To: Ulf.Engstr?m at mb04.swip.net if I use mb04.swip.net as outgoing server,

...instead.

>From RFC822:

     6.  ADDRESS SPECIFICATION     6.1.  SYNTAX
     address     =  mailbox                      ; one addressee
                 /  group                        ; named list
     group       =  phrase ":" [#mailbox] ";"
     mailbox     =  addr-spec                    ; simple address
                 /  phrase route-addr            ; name & addr-spec
     route-addr  =  "<" [route] addr-spec ">"
     route       =  1#("@" domain) ":"           ; path-relative
     addr-spec   =  local-part "@" domain        ; global address
     local-part  =  word *("." word)             ; uninterpreted
                                                 ; case-preserved
     domain      =  sub-domain *("." sub-domain)
     sub-domain  =  domain-ref / domain-literal
     domain-ref  =  atom                         ; symbolic reference

 --------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com
--------------------------------------------




From skaller at maxtal.com.au  Tue Nov 30 09:39:58 1999
From: skaller at maxtal.com.au (skaller)
Date: Wed, 01 Dec 1999 01:39:58 +1100
Subject: Python complaints
References: <1268241794-29392927@hypernet.com> <FLzGo6.AAn@world.std.com>
Message-ID: <3843E1BE.DE6CC8E0@maxtal.com.au>

Will Ware wrote:

> The reasoning leaned heavily on functions
> for the successor and predecessor of an integer. Conceivably there
> might be some value in adding successor and predecessor functions to
> immutable types where they had obvious meanings. Not that I ever do
> number theory myself, nor am I aware of anybody using Python for it,
> so this isn't a wish list thing, this is just idle conjecture.

Succ and pred were the FIRST, the VERY FIRST, methods added
to any viper type....

# This module is exclusive to Viper
# It is a builtin module, defining classes for all
# the Python types.
#
# Methods and attributes for objects can be found in the
# class dictionary of the typing class, even when the object
# is not a PyInstance of the class.
import string
import sys

# datum types
class PyIntType : 
  def succ(x): return x + 1
  def pred(x): return x - 1

.....

mainly because they were easy to implement :-)

-- 
John Skaller, mailto:skaller at maxtal.com.au
10/1 Toxteth Rd Glebe NSW 2037 Australia
homepage: http://www.maxtal.com.au/~skaller
voice: 61-2-9660-0850



From gmcm at hypernet.com  Wed Nov  3 12:11:01 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 3 Nov 1999 12:11:01 -0500
Subject: exceptions from C
In-Reply-To: <Pine.LNX.4.10.9911040950440.25629-100000@crb.crb-web.com>
Message-ID: <1270473439-35788847@hypernet.com>

Linux User wrote:

> I have been reading the section in the Lutz book on exceptions. 
> It seems that exceptions are no longer simply string objects. 
> Does this change the way exception are raised from C?  If so
> where might I find info on how to do this?

bltinmodule.c has pointers to all the standard exception 
classes (initialized during python startup - they're really 
created in pure Python).

To raise one from C, you do something like:
   PyErr_SetString(PyExc_IndexError, "index out of range");
   return 0;

Creating your own exception classes from C will take more 
work. As usual with the C API, how much work depends on 
how clever you are <wink>.

grep-is-your-best-friend-ly y'rs


- Gordon



From josefb at csi.com  Wed Nov 17 07:40:48 1999
From: josefb at csi.com (Josef Betancourt)
Date: Wed, 17 Nov 1999 07:40:48 -0500
Subject: Need help dealing with spaces in NT paths.
References: <80pr76$jq5$1@nnrp1.deja.com> <11f733ec.96958819@usw-ex0101-007.remarq.com> <uogcu93dt.fsf@cs.uu.nl> <80sj4q$l7t$1@nnrp1.deja.com> <80tgi8$hbn$1@ssauraaa-i-1.production.compuserve.com> <002501bf30e8$a41a5bc0$f29b12c2@secret.pythonware.com>
Message-ID: <001301bf30f8$fa61cf80$224dc0d8@josefbet>

Maybe my typing is really off today, but it don't work:

>>> os.system('dir "program files"')

 Volume in drive C is CDRIVE
 Volume Serial Number is 2E50-0FE9
 Directory of C:\WINDOWS

File not found
                      362,876,928 bytes free
0
>>>

----- Original Message ----- 
From: Fredrik Lundh <fredrik at pythonware.com>
To: T. C. Mits <71351.356 at compuserve.com>
Cc: Python Mailing List <python-list at cwi.nl>
Sent: Wednesday, November 17, 1999 5:43 AM
Subject: Re: Need help dealing with spaces in NT paths.


> 
> T. C. Mits <71351.356 at compuserve.com> wrote:
> > A simple example of the difficulty is:
> > 
> > >>> os.system(r'"dir Program Files"')
> > Bad command or file name
> > 0
> > >>>
> 
> strangely enough, 
> 
>     "os.system(r'dir Program Files')"
> 
> doesn't work either.  nor does
> 
>     c:\> "dir program files"
> 
> at the command prompt.
> 
> (when you get this far, you should have
> realised how to fix this...  but alright, I'll
> spell it out:
>    
>     c:\> dir "program files"
>     os.system('dir "program files"')
> 
> but in this specific case, you should use
> "os.listdir" or perhaps "glob.glob" instead,
> but that's another story...)
> 
> </F>
> 
> <!-- (the eff-bot guide to) the standard python library:
> http://www.pythonware.com/people/fredrik/librarybook.htm
> (news: just posted an errata for the 'first printing') -->
> 




From root at anakin.alzzalinux.com  Tue Nov 16 21:21:49 1999
From: root at anakin.alzzalinux.com (root)
Date: Wed, 17 Nov 1999 02:21:49 GMT
Subject: Debian/Apache problems
References: <80o5ub$csn$1@nnrp1.deja.com>
Message-ID: <slrn834443.iis.root@anakin.alzzalinux.com>

><Directory /usr/lib/cgi-bin/python>
>PYTHONPATH /usr/lib/cgi-bin/python/stdlibs
></Directory>

I don't think that prob is related to Python.

I think the right format is like this:

SetEnv PYTHONPATH /usr/lib/cgi-bin/python/stdlibs
--
ALZZA Linux 6.1 (May the Force be with you)
						Man-Yong Lee



From alex at somewhere.around.here  Wed Nov 24 21:33:39 1999
From: alex at somewhere.around.here (alex)
Date: 25 Nov 1999 13:33:39 +1100
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <867lj85g4c.fsf@g.local>
Message-ID: <bvwso1vcnb0.fsf@room.mathdom.anu.edu.au>

> suppose you have a piece of code that builds a big data structure out
> of class instances; then every time you make a small change and run it
> again (you have to run it again, because existing instances belong to
> the old versions of the classes) you can be leaking a lot of memory.

Sometimes you don't have to run the whole thing again.  E.g.

>>> class A: 
...   def __init__(self,data):self.data=data
... 
>>> t=A(5)
>>> t
<__main__.A instance at 815d370>
>>> class A:
...   def __repr__(self):return "This really is the new version"
... 
>>> t.__class__=A
>>> t
This really is the new version
>>> t.data
5
>>> 

This can be very useful if you want to fiddle with a method in an object
that takes a long time to initialize.  

But you're probably right about the memory management.

Alex.



From mwh21 at cam.ac.uk  Sat Nov  6 08:03:25 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 06 Nov 1999 13:03:25 +0000
Subject: Get parameter by name
References: <Pine.LNX.4.20.9911061213410.18974-100000@fep132.fep.ru>
Message-ID: <m366zfai1u.fsf@atrus.jesus.cam.ac.uk>

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

> Hello!
> 
>    I have a function with a long list of arguments:
> 
> def f(a1, a2, a3, b1, b2, b3, td, sd, si, ka, num=0, date=None):
>    ...body_here...
> 
> and I have a list of strings "a1, td, ka". How can I get the value of
> corresponding args? E.g., if I have
> 
>    x = "ka"
> 
> how can I get the value of parameter, the name of which is in x ("ka", in
> this case)?
> 
> Oleg.

locals() should do the trick:

def f(x):
    return locals()["x"]

Arguments are pretty much identical to local variables.

HTH,
Michael



From culliton at clark.net  Thu Nov 11 17:45:03 1999
From: culliton at clark.net (Tom Culliton)
Date: Thu, 11 Nov 1999 22:45:03 GMT
Subject: Beginning programming with Python
References: <sng0ychgwt.fsf@motorola.com> <382B3A53.5816B2@callware.com> <382B3D79.74CD3E0F@email.sps.mot.com>
Message-ID: <PFHW3.36476$oa2.142129@iad-read.news.verio.net>

In article <382B3D79.74CD3E0F at email.sps.mot.com>,
Aaron Ginn (rp2885) <rp2885 at email.sps.mot.com> wrote:
> That's what I expected.  I perused the books in question and musc of
> the syntax looked _exactly_ like Perl.

I'd say it's a lot more sparse and doesn't use anywhere near as many
"punctuation" characters.  "Simpler" in that it has fewer operators,
forms, and implicit actions.

> An honest opinion, please; is Python better suited for beginners
> than Perl, as I've heard?

Absolutely.  Perl has far more, magic, special cases, short cuts, odd
notation, magic variables, ...  More "stuff" to learn.  Python is much
more straightforward because there are simply less details to get lost
in.



From calishar at *NOSPAM*home.com  Mon Nov 29 21:32:41 1999
From: calishar at *NOSPAM*home.com (Calishar)
Date: Tue, 30 Nov 1999 02:32:41 GMT
Subject: WinNT and DLL's
Message-ID: <ddd64s4pp498hb7qlrgnq6o8fdqilrr27d@4ax.com>

Hi Folks,

  I'm in the middle (almost literally, lots of paper) of developing an
application to be used on Windows NT, and maybe Win'95 systems. One of
the things I have not managed to find available anywhere, is an
interface to PkZip/Pkunzip zipped programs. I have tried using the
gzip and zlib modules on the files I am working with, and they don't
like the input files.

  In searching for info about unzipping, I ran across the info-zip
unzip (and zip) source code, however I do not have the knowledge in
c/c++ to implement a python interface to these dll's.

  Is there a way that I can simply access the pre-compiled dll files?
(I noticed a loadlibrary method in the (truly excellent) win32api
module, can I do something with this?)

  Or alternately, is there someone who would be willing and able to
perform whatever magic is required to compile the dll so that Python
can access it?

  I really hate asking for this level of help when I don'thave much
(if anything) to offer in return, but a search on Deja, and in the FAQ
didn't point me to any easy answers) and I have tried reading the swig
docs, but my brain just hit snooze, and never woke up till the next
day. (Not a knock about the doc writer, just a comment about what I
can read and comprehend)

Calishar



From mhammond at skippinet.com.au  Sun Nov 14 23:02:09 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Mon, 15 Nov 1999 04:02:09 GMT
Subject: NT admin scripts with python
References: <000201bf2e34$26d93750$020a10ac@p350.gnaa.net> <80lhgl$22gn$1@hub.org> <80ncp9$rnm$1@nnrp1.deja.com>
Message-ID: <5BLX3.4099$Xo.22270@news-server.bigpond.net.au>

damien_morton at my-deja.com wrote in message <80ncp9$rnm$1 at nnrp1.deja.com>...
>For doing NT admin scripts, I have found that System Scripting Host
>from www.netal.com is extremely usefull.
>
>You can happily program it in Python, VBscript, Perl or whatever. Its
>basically a framework for running scripts as NT services.

I had a quick look at this, and it does indeed seem a cool little piece of
software.  If you do try and use it with Python, make sure you are running
win32all-127 - 125 and earlier had a bug that showed itself with this
tool...

Mark.






From exeq at my-deja.com  Thu Nov  4 08:04:29 1999
From: exeq at my-deja.com (exeq at my-deja.com)
Date: Thu, 04 Nov 1999 13:04:29 GMT
Subject: Tkinter and Solaris 7 Sparc [Q]
References: <7vqblc$ham$1@nnrp1.deja.com> <3820F405.4A1C8767@python.net>
Message-ID: <7vs08s$ll1$1@nnrp1.deja.com>

In article <3820F405.4A1C8767 at python.net>,
  "Thomas A. Bryan" <tbryan at python.net> wrote:
> exeq at my-deja.com wrote:
> > I am trying to compile Python 1.5.2 with tkinter on Solaris 7
(Sparc).
>
> A word of encouragement: I have successfully built Python 1.5.2 on
> Solaris 7.
> I can't remember whether I used gcc or Sun's compiler.
>
> > have installed TCL/TK 8.0.5, GCC 2.8.1 / Sun Workshop 5 (I tryied
with
> > both compilers).
> >
> > I just get an error saying that libX11.so.6.1 is needed. Is there a
way
> > to build tkinter without X11? Or I just need to download and install
> > X11?
>
> First, I ask the "obvious" question.  Did you edit the Modules/Setup
> file
> (or, better, did you make a Setup.local file)?  If so, did you make
sure
> to use the entries in the Tkinter portion that are specific to
Solaris?
>
> (from Setup)
>
> # *** Uncomment and edit to reflect where your X11 header files are:
> #-I/usr/X11R6/include \
> # *** Or uncomment this for Solaris:
>        -I/usr/openwin/include \
> .....
> # *** Uncomment and edit to reflect where your X11 libraries are:
> #-L/usr/X11R6/lib \
> # *** Or uncomment this for Solaris:
>        -L/usr/openwin/lib \
>
> If you already did that correctly, is libX11 in that directory?  That
is
> use 'ls' and 'find' to look for the library.  Is it there?
>
> ---Tom
>

Hi

Yes, I did that, but my error was that I used Tcl8.0.5 and tk from the
Metalab repository (binaries pre-compiled). And it was compiled against
X11R6 libs. So I downloaded the sources and compiled myself. Everything
worked fine.

I tryied also to used Tcl/tk 8.2 and it didn't work. The error stated
that: -ltcl8.2  -ltk8.2 : not found. However the libtcl8.2.so and
libtk8.2.so where in the LD_LIBRARY_PATH and in PATH ... :(

Anyway, I am using tcl8.0.5 and it works.

Thanks to all that replied.

/B


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



From greg.ewing at compaq.com  Tue Nov 16 10:25:59 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Wed, 17 Nov 1999 04:25:59 +1300
Subject: Beginning programming with Python
References: <sng0ychgwt.fsf@motorola.com> <382B3A53.5816B2@callware.com> <382B3D79.74CD3E0F@email.sps.mot.com> <80p9ds$ppo$1@newshost.accu.uu.nl> <asteroidal-80psfo/INN-2.2.1/bat@broadway.news.is-europe.net>
Message-ID: <38317787.19AFDAF5@compaq.com>

Janos Blazi wrote:
> 
> the indenting this time? There is actually nothing else you can forget in
> Python.

Well, not quite - you can forget the "self" in method
declarations fairly easily...

Greg



From aahz at netcom.com  Mon Nov  8 18:11:50 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 8 Nov 1999 23:11:50 GMT
Subject: [Python] header (fwd)
References: <Pine.HPX.4.20.9911081711570.27637-100000@ztt.rsn.hp.com>
Message-ID: <807lbm$ail$1@nntp5.atl.mindspring.net>

In article <Pine.HPX.4.20.9911081711570.27637-100000 at ztt.rsn.hp.com>,
Benoit Ciceron  <bciceron at rsn.hp.com> wrote:
>
>can we get an automatic subject header : "[Python]" to ALL messages
>submitted to the list , including replies and newsgroup threads ?

Why?
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From aahz at netcom.com  Wed Nov 10 12:16:51 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 10 Nov 1999 17:16:51 GMT
Subject: Intelligent strptime?
References: <19991110171256.A5196@optiplex.palga.uucp>
Message-ID: <80c9a3$r8$1@nntp9.atl.mindspring.net>

In article <19991110171256.A5196 at optiplex.palga.uucp>,
Gerrit Holl  <gerrit at nl.linux.org> wrote:
>
>Does there exist an intelligent strptime? "26/11/99", for example, can _only_
>mean the 26th November of this year, and the chance that "14:11:12" means
>14 november 1912 is very small. Is there a module/function that guesses the
>format? There isn't doubt very often, only for 12/11/99, US is 11 december,
>Europe is 12 November (isn't it?).

I don't have my docs handy, but I think rfc822 and/or mxDateTime might
have what you want.
--
> I wrote:
> > a slightly hairer design issue is what combinations
> > of pattern and string the new 're' will handle.
> 



From robert.meegan at murl.com  Wed Nov 10 16:19:55 1999
From: robert.meegan at murl.com (Robert Meegan)
Date: Wed, 10 Nov 1999 21:19:55 GMT
Subject: random function
In-Reply-To: <3829D78B.7FC142F4@scripps.edu>
References: <3829D78B.7FC142F4@scripps.edu>
Message-ID: <19991110.21195500@robert.polytopic.com>

Olivier -

The is both a random and a whrandom module included with the Python 
distribution. Both are identified in the Python Library Reference 
index.

- Robert

>>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<<

On 11/10/99, 2:37:32 PM, Olivier Roche <olivier at scripps.edu> wrote 
regarding random function:


> Hi everybody

> By looking at the index of the python documentation index i haven't 
seen
> any
> random function description. Could somebody tell me if a random 
function

> to generate random number (between 0 and 1) exists ?

> Thanks,

> Olivier


> --
> http://www.python.org/mailman/listinfo/python-list






From pinard at iro.umontreal.ca  Mon Nov 29 17:11:51 1999
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 29 Nov 1999 17:11:51 -0500
Subject: '==' vs. 'is' behavior
In-Reply-To: duncan@rcp.co.uk's message of "15 Nov 1999 13:08:14 GMT"
References: <382C2C6A.64A8F2C2@theriver.com> <14380.12036.197083.409565@weyr.cnri.reston.va.us> <8E7F86267duncanrcpcouk@news.rmplc.co.uk>
Message-ID: <oqemd9dk2g.fsf@titan.progiciels-bpi.ca>

duncan at rcp.co.uk (Duncan Booth) writes:

> >>> print a is b

By the way, is there common wisdom (or rather, maybe, usage-standards-to-be)
about using `is' instead of `=='?

In most cases I have seen so far, it is unusual that one really has to
resort to `is', because `==' usually does something proper on average.
But even then, I prefer writing `is None' than `== None', because it is
more exact.  I could of course use the tested things as mere logical values,
as a way to escape that tiny problem (:-), but sometimes it is just clearer
to write things explicitly.

So, I was just wondering if someone would not have a clear set of good
habits on that matter.  (I feel just a bit like "M?g?re", who indirectly
induced the Trojan war by offering the golden apple to three goddesses :-).

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





From AKiss at GEOMETRIA.hu  Mon Nov 15 06:01:11 1999
From: AKiss at GEOMETRIA.hu (Kiss, Arpad)
Date: Mon, 15 Nov 1999 12:01:11 +0100
Subject: win32api... another question:  GetEnvironmentVariable("PATH")
Message-ID: <01BF2F61.1D2B8DB0@exchange1.geometria.hu>

Hi,

Kevin Cazabon <kcazabon at home.com> wrote in message 
news:<fcMX3.65012$Rx2.773564 at news1.rdc1.bc.home.com>...
> I can't seem to get the win32api.GetEnvironmentVariable() function to 
work
> properly... am I not using it correctly by the following?
> 
> path = win32api.GetEnvironmentVariable("PATH")

It works for me on a Windows NT Workstation (sp5).

> 
> or
> 
> path = win32api.GetEnvironmentVariable("%%PATH%%")

It doesn't work....





From gmcm at hypernet.com  Tue Nov  9 11:05:16 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 9 Nov 1999 11:05:16 -0500
Subject: [Python] header (fwd)
In-Reply-To: <209f01bf2ac7$c19b1d50$f0c809c0@lslp7o.lsl.co.uk>
Message-ID: <1269958983-19391623@hypernet.com>

Ton J. Ibbs writes:

> Robert Meegan <rm at murl.com> wrote (in answer to Benoit Ciceron):
> >If what you want is a way to filter mailing list messages, may I
> >suggest that you look for "python-list" in both the "To:" and
> "CC:" >fields.
> 
> Unfortunately, demonstrably insufficient (i.e., some messages to
> the list do not actually have python-list in either To: or CC:
> fields). This is a problem to those (for instance) who are
> unlucky enough to have to use Outlook Express (<fx
> spitting>phthwip-ping</>) which only allows filtering on To,
> From, CC and Subject fields. Personally, I just put up with some
> messages getting misfiled to my inbox, and dream of one day
> getting back to a decent mail handler...

Huh? I filter (in Pegasus mail) on "python-list@" in To: and 
CC: and get them all.

- Gordon



From mhammond at skippinet.com.au  Sun Nov 14 23:35:36 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Mon, 15 Nov 1999 15:35:36 +1100
Subject: NT admin scripts with python
In-Reply-To: <002d01bf2f22$34a16850$020a10ac@p350.gnaa.net>
Message-ID: <00d001bf2f22$dd440800$0501a8c0@bobcat>

> Win32::Console: Perl module for Windows console interaction
We dont have this.

> Win32::EventLog: Read from and write to the Windows NT event log
> Win32::File: Manage file attributes (read only, system, hidden...)
> Win32::FileSecurity: manage ACLs in Perl

All of the above - win32evtlog, win32file, win32security.

> Win32::Internet: Perl module for Internet extensions

win32com.internet

> Win32::NetAdmin: Administer users and groups
> Win32::NetResource: Manage resources (servers, file shares,
printers)

Both in win32net

> Win32::Process: Start and stop Win32 processes

win32process

> Win32::Registry: Read and manage the Win32 Registry

win32api

> Win32::Service: Manage Windows NT services

win32service (and we can even implement native services in Python)

There are also plenty more available for Python, but I have only
responded to the ones listed.

Mark.




From mlh at vier.idi.ntnu.no  Fri Nov 26 11:43:36 1999
From: mlh at vier.idi.ntnu.no (Magnus L. Hetland)
Date: 26 Nov 1999 17:43:36 +0100
Subject: Installing NumPy in a non-standard place...
Message-ID: <y0jhfi9xkxz.fsf@vier.idi.ntnu.no>

Installing Python is a dream - but NumPy seems to be a bit stubborn...
How can I tell it where I want to install it? I have copied
Makefile.pre.in from my python installation, but still it wants to
install it in a standard location (that I don't have access to,
because I don't have root access on the machine)... Should I just copy
all the files manually? Am I missing something? (Couldn't see anything
in the installation instructions...)

--

  Magnus          Echelon jamming noise:
  Lie             FBI CIA NSA Handgun Assault Bomb Drug Terrorism
  Hetland         Special Forces Delta Force AK47 Hillary Clinton 



From fdrake at acm.org  Wed Nov 24 09:50:26 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Wed, 24 Nov 1999 09:50:26 -0500 (EST)
Subject: Python complaints
In-Reply-To: <1268673769-3409210@hypernet.com>
References: <867lj85g4c.fsf@g.local>
	<1268673769-3409210@hypernet.com>
Message-ID: <14395.64306.499503.83970@weyr.cnri.reston.va.us>

Gordon McMillan writes:
 > >>> def p(score):
 > ...  print "You scored %d point%s" % (score, score==1 and ' ' 
 > or 's')

  The catch is that this only works when the result value for the true 
case is also true.  I often find I want this sort of construct when
the true result is either false or may be either.  The hack to get
around this:

       x = (some.test and [v1] or [v2])[0]

is really nasty in practice, so I find it's typically *far* more
reliable to expand it out to a full-fledged if:

       if some.test:
           x = v1
       else:
           x = v2

  Much longer, but more reliable given the lack of ?: (which I'd also
like to see added, though I'm not sure what the syntax should be).


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From mlh at vier.idi.ntnu.no  Tue Nov  9 08:41:40 1999
From: mlh at vier.idi.ntnu.no (Magnus L. Hetland)
Date: 09 Nov 1999 14:41:40 +0100
Subject: A Modest Proposal -- <wink>
References: <99110613580407.02970@quadra.teleo.net> <8072o6$m4i$1@newshost.accu.uu.nl>
Message-ID: <y0jiu3b2357.fsf@vier.idi.ntnu.no>

m.faassen at vet.uu.nl (Martijn Faassen) writes:

> 
> As you can see, XWTSNTSDL is fully consistent with your current proposals. 
> It is therefore expected the major internet browsers will soon be able to
> support these new standards.

It is not consistent with XML... It would have to be <wink/>....

> 
> Regards,
> 
> Martijn Faassen, PSU (Python Secret Underground)

--

  Magnus          Echelon jamming noise:
  Lie             FBI CIA NSA Handgun Assault Bomb Drug Terrorism
  Hetland         Special Forces Delta Force AK47 Hillary Clinton 



From freshhorses at earthlink.net  Mon Nov 29 14:08:56 1999
From: freshhorses at earthlink.net (freshhorses at earthlink.net)
Date: Mon, 29 Nov 1999 19:08:56 GMT
Subject: Embedded size?
Message-ID: <3842cf05.407909516@news.earthlink.net>

Were I to embed Python as my in-application scripting language, how
much space would it need in both code and memory?  I don't see
anything obvious on the org site discussing those sorts of
requirements :)






From conversy at lri.fr  Wed Nov 17 11:50:15 1999
From: conversy at lri.fr (Stephane Conversy)
Date: Wed, 17 Nov 1999 17:50:15 +0100
Subject: What does Tkinter do when Idle ?
Message-ID: <3832DCC6.60E05050@lri.fr>

It seems that from time to time (say every 10 s.) , Tkinter wakes up
and do something, and then comes back to idle.

is it right ?

    stef

--
St?phane Conversy
http://www-ihm.lri.fr/~conversy/
mailto:conversy at lri.fr






From malcolmt at smart.net.au  Sun Nov  7 22:37:40 1999
From: malcolmt at smart.net.au (Malcolm Tredinnick)
Date: Mon, 8 Nov 1999 14:37:40 +1100
Subject: Graph Layouts (revisited)
Message-ID: <19991108143740.A895@Ridcully.home>

OK, I was bored on the weekend (well, it was raining on Saturday and
playing on the computer seemed more fun than cleaning the apartment). So I
decided to play around with some graph layout algorithms as a response to
the recent thread where somebody requested such a monster.

Not quite at the point where I'm going to release code yet, but I have a
couple of questions which souls more knowledgeable in graph theory than I
might be able to answer. If people want to just reply via direct email,
rather than to the list, that is fine by me (since the discussion may get a
little dry and technical). However, sometimes I find others' tangential
discussions on here to be interesting anyway, so you decide.

So far I have implemented an algorithm (in Python, of course) that lays out
undirected graphs in a reasonably visually pleasing manner (following an
algorithm  from the proceedings of the DIMACS International Workshop on
Graph Drawing (held at Princeton in 1994 -- proceedings are are the
Springer Lecture Notes in Computer Science, vol 894). My question concerns
*directed* graphs: what extra layout criteria are usually used here? I
assume it would be nice to have as many arrows as possible pointing in the
same general direction -- so that flows in the graph move generally, say,
from left to right.  Anything else that is "usual"?

Just to give people a flavour for what I have done so far: I am leaving the
issues of how to get the graph data in the first place and how to display
the laid out graph at the end up to the user. I am simply writing a class
that is given a graph (which is just a dictionary containing the number of
vertices in the graph and a list of two-tuples (ordered for directed
graphs) which represent the edges) and generates a list of coordinates for
placing the various vertices. Of course, debugging such an abstract class
is a little interesting on its own, so I have also written a module that
generates various "standard" graphs (e.g. complete graphs, chains, wheels,
square and triangular grids, ...) for passing to the algorithm class. I
also have a nearly complete back end that outputs encapsulated postscript.
My intention is to try and write some interfaces to PIDDLE if I feel so
motivated.

Problems that need to be dealt with: The algorithm assumes that all nodes
are basically of zero size. That is, it doesn't worry about wrapping edges
around nodes of significant size. This is on the "to do" list, but is
definitely a version 2 feature.

Well, as usual, I have rabitted on and on and this reads like a vapourware
announcement. The important bit is paragraph 3. Can you help?

Thanks in advance,
Malcolm Tredinnick



From ivanlan at callware.com  Mon Nov 15 13:18:32 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Mon, 15 Nov 1999 11:18:32 -0700
Subject: Directory chooser in Tk
Message-ID: <38304E78.F92C8596@callware.com>

Hi All--
Yet another dumb question about Tkinter.

I need a dialog similar to the askopenfilename() dialog, but for
directories.  Is there such a beast?  Or are there options to pass to
askopenfilename() that give me the same result?

I don't want to use win32 specific calls if I can avoid it.

<everything-i-want-doesn't-exist-i-know-it>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours
----------------------------------------------



From mikael.johansson at helsinki.xx  Thu Nov 11 09:33:31 1999
From: mikael.johansson at helsinki.xx (Mikael Johansson)
Date: Thu, 11 Nov 1999 16:33:31 +0200
Subject: print "foo", without a space
Message-ID: <382AD3BB.69899CBD@helsinki.xx>

Hello All!

Is there another way of getting print to not make a newline than
using:
    print "foo",  ?
This because the comma always produces a space which necessary
isn't that nice, for example:

for i in range (1,30):
    print i,
    if i%10==1 and i<>11: print "st",
    elif i%10==2 and i<>12: print "nd",
    elif i%10==3 and i<>13: print "rd",
    else: print "th",
    print "fish"
 
Now you would like the "endings" to come right after the number.
There is naturally always some way to get around this, in this
case for example:

end=["st","nd","rd","th"]
for i in range(1,30):
    print `i`+end[max(min(abs(i)%10,4)-1-
    4*(not(int('0'+`abs(i)`[-2:-1])-1)),-1)],'fish'

OK, this got a bit out of hand, the above looks terrible, does 
anyone see how it could be simplified? At least it's more general
than the above as it can handle the whole range of integers,
including negative ones :-) However it shows that it would be
less obfuscating to have a print statement that could continue
printing on the same line without a space.

If there isn't an easy way for this, I suggest the following:
    print "foo".
So that a dot at the end of the line would mean "no space either".

Appreciating any comments.

Have a nice day,
    Mikael J.
    for e-mail note that the country code for Finland is not xx



From ivanlan at callware.com  Fri Nov 12 09:53:11 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Fri, 12 Nov 1999 07:53:11 -0700
Subject: Tkinter PhotoImage Question
Message-ID: <382C29D7.DBA3E66@callware.com>

Hi All--

[I sent this Wednesday but haven't seen it appear on the list, so I
assume the orginal is lost in the solar flares somewhere.  If it turns
up twice, I apologize.]

Ok, I've managed to create an image, using xx=PhotoImage(file), and I
can retrieve pixel values from that image using xx.get(x,y).

Now I want to use the put() method to write a pixel.

How do I do that?  I would have expected that put() would take a string
of pixel values for the data argument, but that doesn't seem to be the
case. ...

And what's the bbox argument?

<there's-a-cat-on-my-screen>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours
----------------------------------------------



From gmcm at hypernet.com  Wed Nov 10 11:12:15 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 10 Nov 1999 11:12:15 -0500
Subject: Call a function in DLL with string arguments?
In-Reply-To: <80bv6i$kp$1@usenet01.srv.cis.pitt.edu>
Message-ID: <1269872166-298367@hypernet.com>

Juping  Jin writes:

> I wrote a DLL and want its functions being available from python.
> Its function accept string arguments. From python, I use windll
> to load the module like:
> 
>  handle = windll.module("dllname")
>  handle.fcnname(arguments)
> 
> Python gives following msg:
> 
> TypeError: illegal argument type for built-in operation
> 
> So I did some tests, wrote a small dll and found out that if the
> function accept only numerical values (actually, only integer
> value produces expected results), then python won't give above
> errors. For char or char * argument, it always gave above msg.
> Did I do something wrong?

Well, if you're serious about exposing it to Python, you should 
probably rewrite it, or wrap it, or SWIG it and make a proper 
extension module. If you want to poke at it with windll, then 
you need to realize that Python objects, even if they have a 
natural translation to a C representation, are completely 
incomprehensible to non-Python aware C. As I recall, there's a 
cstring object in the same download as windll which gives you 
a buffer you can use to communicate with a function in the dll. 
In particular, you initialize the cstring with your Python string, 
and it takes care of getting the function a real char *.

Sorry to be vague, but I no longer have windll etc. installed on 
my system.

- Gordon



From cjc26 at nospam.cornell.edu  Sun Nov  7 20:29:34 1999
From: cjc26 at nospam.cornell.edu (Cliff Crawford)
Date: 8 Nov 1999 01:29:34 GMT
Subject: __str__ vs. __repr__
References: <000001bf275c$02370d20$1d2d153f@tim>
Message-ID: <slrn82c9qk.qs7.cjc26@tankgrrl.bridget.mindriot.net>

I'm glad this thread cropped up when it did.  I've been writing a
simple parser module, and had been trying to track down a "bug"
involving the token scanner inserting extra backslashes in strings.
But then I thought of the __repr__ vs __str__ distinction, and
tried this:

>>> r'blah\\blah'
'blah\\\\blah'
>>> 'blah\\blah'
'blah\\blah'
>>> print r'blah\\blah'
blah\\blah
>>> print 'blah\\blah'
blah\blah

So actually there wasn't anything wrong with my code, it was just
that I was printing my results in the interpreter, which uses
__repr__ and thus inserts an extra backslash for each backslash in
the string.
Well, it only cost me about a half hour of needless debugging =)


-- 
cliff crawford   http://www.people.cornell.edu/pages/cjc26/
-><-                        Shall she hear the lion's roar?



From tbryan at python.net  Tue Nov  2 21:32:21 1999
From: tbryan at python.net (Thomas A. Bryan)
Date: Tue, 02 Nov 1999 21:32:21 -0500
Subject: Comparison between Python and "Ruby"
References: <1270663667-24391282@hypernet.com> <381ED7A6.D7D7FB13@compaq.com> <877lk1wiij.fsf@ev.netlab.co.jp> <19991102154543.A4469@optiplex.palga.uucp>
Message-ID: <381F9EB4.63FA033F@python.net>

Gerrit Holl wrote:
> 
> Yukihiro Matsumoto wrote:
> >
> > I guess he loves reducing variablity, if possible, which is totally
> > against TMTOWTDI.  I don't judge whether it is good or bad.
> >
> 
> Are there people who see TMTOWTDI [1] as an advantage?
> [1]
> There's more than one way to do it.

Yes, there are.  Larry Wall's speeches (Wall is the Guido of Perl)
often proclaim that the problem with many programming languages 
is that they enforce too many restrictions on the programmer and 
his code.  Thus, given a problem, the programmer first has to map 
that problem onto what is permitted by the language.  The less 
flexibility the language offers, the more difficult this mapping 
process can be for more problems.  Perl is supposed to reduce the 
amount of mapping that a programmer has to do before implementation.

It's an interesting concept, but I don't know whether I agree that 
this aspect of Perl is a "feature."  I haven't been programming long 
enough to decide.  Some experienced C/C++ software engineers that 
I've known complain that Larry Wall is ignoring the years of industry 
and academic wisdom in software engineering when he says things like 
that.  Code typically spends a lot of time in maintenance by a lot 
of different people.  (If it doesn't, it simply dies.)  If a language 
offers too much flexibility, then it doesn't matter if you know the 
language.  You still may have a really hard time understanding someone 
else's code in that language.  If you think about it, a lot of people's 
Python code looks very similar in structure and syntax.  I don't think 
that's as true in the Perl world.  

In my own personal experience, I've never had trouble reading other
people's Python code.  I often have to struggle through other people's
Perl code.  Even when I knew more Perl than Python, this statement was 
still true.  When I read Perl code, I spent too much of the time 
figuring out *how* the code worked instead of figuring out *what* the 
code was doing.  I suspect that this difficulty was due in part to 
the fact that Perl has more Perl hackers than Python has Pythonistas.
Thus, it also has more *bad* programmers producing *bad* code.

To keep myself from rambling too much, I'll summarize by simply saying 
that if Python is more restrictive than Perl, I've never noticed it, and 
I've never felt that my options were limited in Python.  On the other 
hand, when I go to the more flexible world of Perl, I seem to spend a 
lot of time just dealing with the language.  Python gives me sufficient 
power and flexibility but then stays out of my way and lets me think 
about data structures, algorithms, and design patterns.  

I'll close with something Tim Peters posted once about "The Python Way."
Note the item about "only one obvious way."  There might be other ways, 
but they're usually non-obvious and not often used in Python.

"""Beautiful is better than ugly.
Explicit is better than implicit.
Simple is better than complex.
Complex is better than complicated.
Flat is better than nested.
Sparse is better than dense.
Readability counts.
Special cases aren't special enough to break the rules.
Although practicality beats purity.
Errors should never pass silently.
Unless explicitly silenced.
In the face of ambiguity, refuse the temptation to guess.
There should be one-- and preferably only one --obvious way to do it.
Although that way may not be obvious at first unless you're Dutch.     
Now is better than never.
Although never is often better than *right* now.
If the implementation is hard to explain, it's a bad idea.
If the implementation is easy to explain, it may be a good idea.     
Namespaces are one honking great idea -- let's do more of those! """

---Tom



From fbi at boxer.no  Tue Nov  9 10:12:31 1999
From: fbi at boxer.no (Frank Biringer)
Date: Tue, 09 Nov 1999 16:12:31 +0100
Subject: Another bloody newbie question: Executable Python files?
References: <382821f1.96924220@212.225.36.99>
Message-ID: <382839DF.8A6E359C@boxer.no>

One possibility is the freeze-module (not shipped with Windows-version; in
case you're using Win32 check the reeze.py from the Python source). Search
www.python.org for 'freeze'.
The module is a little bit tricky and you'll need a c-compiler.

Endgamer wrote:

> Is it possible (I know it's not required) to compile a Python program
> to be stand-alone? Or is it required that Python be installed on every
> machine the program runs on? I'm working with Windows machines, and
> I'm writing this little script I need to have run on each machine (of
> about 20) at startup - I'd prefer not to have to install Python on
> each machine!
>
> Thanks in advance...
>
> Endgamer
> Spider Warrior
>
> (Endgamer71 at spamtrap.hotmail.com)
> #####################################
> Not a quote, just an on-the-spot observation:
> You know you're a geek when your PC costs exactly
> 15 times as much as your stereo. And gives better
> sound quality.<sigh>
>                               Endgamer
> #####################################

--
Hilsen,
_________
 -|- rank Biringer.

______________________________________________________
Systemanalytiker                 Boxer Technologies AS
tel: (+47) 55 19 20 34           mailto:fbi at boxer.no
fax: (+47) 38 01 90 01           http://www.boxer.no/
______________________________________________________





From prestonlanders at my-deja.com  Thu Nov 18 20:54:12 1999
From: prestonlanders at my-deja.com (Preston Landers)
Date: Fri, 19 Nov 1999 01:54:12 GMT
Subject: help w/assigning lists to dict
References: <8125b9$kuf$1@nnrp1.deja.com>
Message-ID: <812ak3$ohb$1@nnrp1.deja.com>

Greetings,

I don't know what the 'official' way is, but this works for me:

my_list = ['key1','val1','key2','val2']
my_dict = {}

for key in range(0, len(my_list), 2):
    my_dict[ my_list[ key ] ] = my_list[ key + 1 ]

There may be some function call to accomplish same, but this seems like
the most direct approach.  AFAIK there is no built in semantics
equivelent to Perl's assign-list-to-hash.

---Preston Landers


In article <8125b9$kuf$1 at nnrp1.deja.com>,
  nielsenjf at my-deja.com wrote:
>
>
> Hello there, this is hopefully a simple question to answer.
>
> If I have, a list of key,value pairs, is there any way to just assign
it
> to a dictionary? I'm thinking of the case in perl, where you can
assign
> a list to an assoc array.
>
> For example:
>
> a=('key1','val1','key2','val2')
> b={ a } ???????????
>
> Is there a common python idiom for this?

--
|| Preston Landers <prestonlanders at my-deja.com> ||


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



From just at letterror.com  Thu Nov  4 06:47:09 1999
From: just at letterror.com (Just van Rossum)
Date: Thu, 4 Nov 1999 12:47:09 +0100
Subject: Exceptions from callbacks
In-Reply-To: <19991104062722.A184112@vislab.epa.gov>
References: <3820DE5B.A5500330@lords.com> <381CA8E5.5D69E64@lords.com>
 <7virj4$b7b$1@towncrier.cc.monash.edu.au> <381FAE6D.73F426FE@lords.com>
 <19991103071727.A87523@vislab.epa.gov> <3820DE5B.A5500330@lords.com>
Message-ID: <l03102804b447229d208b@[193.78.237.174]>

At 6:27 AM -0500 11/4/99, Randall Hopper wrote:
>So you need the C equivalent of:
>
>      traceback.print_exc()
>
>I don't know what this is, but hopefully someone else will chime in.

It's PyErr_Print().

Just





From phd at phd.russ.ru  Fri Nov 19 03:51:11 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Fri, 19 Nov 1999 08:51:11 +0000 (GMT)
Subject: Python in DOS
In-Reply-To: <19991118185812.A949@better.net>
Message-ID: <Pine.LNX.4.21.9911190850450.12519-100000@fep132.fep.ru>

On Thu, 18 Nov 1999, William Park wrote:
> On Thu, Nov 18, 1999 at 06:24:25PM -0500, Curtis, Craig M. wrote:
> > Does anyone know the "approved" way to get rid of the following message:
> > 
> > Could not find platform dependent libraries <exec_prefix>
> > Consider setting $PYTHONHOME to <prefix>[:<exec_prefix>]
> > 'import site' failed; use -v for traceback
> > Python 1.5.2 (#6, Nov 18 1999, 17:19:05)  [GCC 2.8.1] on ms-dos7
> > Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> > 
> > 
> > I've tried everything I can think of to set PYTHONHOME and PYTHONPATH and
> > python ALWAYS starts up this way.
> 
> If I remember (this was long long time ago), I had to remove case
> sensitivity for library files by setting some dummy environment
> variable.

set LFN=n
   perhaps

Oleg.
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From a.eyre at optichrome.com  Wed Nov  3 12:01:41 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Wed, 3 Nov 1999 17:01:41 -0000
Subject: Please Help!!!
In-Reply-To: <19991103144642.VMTQ11176.fep03-svc@fep05-svc.tin.it>
Message-ID: <002901bf261d$1994c330$3acbd9c2@peridot.optichrome.com>

> Text..., i would build a Tnet-Client similar the Linux console...
> I don't want to use the M$ telnet client :(

Okay. Win32/text based...
I assume you're after a non-line buffered input.

Try this. It's not quite right, but you can fiddle...

# Dodgy telnet client for win32
# Portions of code snipped from Python-1.5.2/Demo/sockets/telnet.py
import msvcrt, socket, select, sys

IAC  = chr(255)	# Interpret as command
DONT = chr(254)
DO   = chr(253)
WONT = chr(252)
WILL = chr(251)

s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
s.connect("myhost", 23)
iac = 0
opt = ""
while 1:
	rl, wl, xl = [s],[],[]
	rl, wl, xl = select.select(rl, wl, xl, 0.1)
	if rl:
		c = s.recv(1)
		if not c:
			break
		cleandata = ""
		if opt:
			print ord(c)
			s.send(opt + c)
			opt = ''
		elif iac:
			iac = 0
			if c == IAC:
				cleandata = cleandata + c
			elif c in (DO, DONT):
				if c == DO: print '(DO)',
				else: print '(DONT)',
				opt = IAC + WONT
			elif c in (WILL, WONT):
				if c == WILL: print '(WILL)',
				else: print '(WONT)',
				opt = IAC + DONT
			else:
				print '(command)', ord(c)
		elif c == IAC:
			iac = 1
			print '(IAC)',
		else:
			cleandata = cleandata + c
		sys.stdout.write(c)
	if msvcrt.kbhit():
		c = msvcrt.getch()
		s.send(c)
		sys.stdout.write(c)

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From mal at lemburg.com  Thu Nov 25 06:18:59 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Thu, 25 Nov 1999 12:18:59 +0100
Subject: mysql and DATE format
References: <81hhm6$ba2$1@nnrp1.deja.com>
Message-ID: <383D1B23.37969F6E@lemburg.com>

sp00fD wrote:
> 
> I've currently got some perl code that executes a sql statment which
> looks like:
> 
> SELECT id, subject, author, text, DATE_FORMAT(date, \"%W, %d %b %Y\"),
> DATE_FORMAT(date, \"%h:%i %p\") FROM Table WHERE thread=18 ORDER by
> date desc
> 
> When trying to use that in python like this:
> """SELECT .. DATE... FROM %s WHERE thread=%d""" % (table, id)
> 
> it tries to expand the DATE %W.. variables.
> I've also tried it as
> 
> sql = 'SELECT id, subject, author, text, '
> sql = sql + 'DATE_FORMAT(date, \"%W, %d %b %Y\"), '
> sql = sql + 'DATE_FORMAT(date, \"%h:%i %p\") FROM' + table
> sql = sql + 'WHERE thread=' + id + 'ORD...'
> 
> which gives me a TypeError: illegal argument type for built-in
> operation.  Why?  How can I do this properly?

Best is to use bound variables and raw strings:

cursor.execute(r'select ... DATE_FORMAT(date,"%W, %d %b %Y")...'
	       r'where thread=?',
	       (id,))

Using bound variables has the advantage of enhancing performance
when multiple executes are done using the same command string
(well at least if the database interface supports this, mxODBC does,
BTW).

-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                    36 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/




From matz at netlab.co.jp  Wed Nov  3 21:20:43 1999
From: matz at netlab.co.jp (Yukihiro Matsumoto)
Date: 04 Nov 1999 11:20:43 +0900
Subject: A Comparison of Python and Ruby Extension Interfaces (Re: Comparison between Python and "Ruby")
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp>
 <7vjtrc$hmm$1@newshost.accu.uu.nl> <38204614.2EAC6051@compaq.com>
 <021119992327527678%pereira@research.att.com>
Message-ID: <87so2nuhdw.fsf@ev.netlab.co.jp>

Fernando Pereira <pereira at research.att.com> writes:

|> * Since Ruby has mark-and-sweep garbage collection, there
|> is no manipulation of reference counts. There weren't
|> actually any examples of this in the Python dbm module,
|> though.
|What's the effect of this on portability? For instance, how are
|pointers to live objects from the C stack handled? As is well-known,
|this cannot be done portably.

See Greg Ewing's article at <38215CE7.70C3D560 at compaq.com>.
In summary, Ruby is fairly portable, but if your first
priority is the portability, Python wins.

						matz.



From kiyolee*remove* at ctimail.com  Sat Nov 27 06:35:28 1999
From: kiyolee*remove* at ctimail.com (Kiyo Kelvin Lee)
Date: Sat, 27 Nov 1999 22:35:28 +1100
Subject: Binary vs non-binary
References: <19991127103504.A1536@stopcontact.palga.uucp>
Message-ID: <81ofle$ipg$1@newton.pacific.net.au>

Binary file is only applicable in DOS/Windows. File opened in binary mode
will treat CR and LF as 2 separate characters while in text mode CR and LF
will be translated into a single CR. It's an old "feature" of C compilers
for DOS/Windows.
Kiyo

"Gerrit Holl" <gerrit.holl at pobox.com> wrote in message
news:19991127103504.A1536 at stopcontact.palga.uucp...
> Hello,
>
> can someone explain me the difference in opening a file in binary mode
> and opening it not in binary mode? And when is some file defined to be
> binary, when it contains an ASCII 0? If so, I understand you need to say
> it's binary in C, but if Python doesn't store it strings with an ASCII 0
> on the end, why is there any difference?
>
> regards,
> Gerrit.
>
> --
> "The world is beating a path to our door"
>
>   -- Bruce Perens, (Open Sources, 1999 O'Reilly and Associates)
>  10:33am  up 17 min,  9 users,  load average: 1.81, 1.47, 0.79
>
>




From x at x.x  Fri Nov 26 22:52:32 1999
From: x at x.x (Eric Jacobs)
Date: Fri, 26 Nov 1999 22:52:32 -0500
Subject: Python complaints
References: <Pine.WNT.4.04.9911261342240.198-100000@rigoletto.ski.org>
Message-ID: <383F5580.49F2D5C9@x.x>

David Ascher wrote:
> 
> Without dealing with the last paragraph, you can fake it now, with
> something like (barely tested):
> 
> import types
> class XRangeClass:
>     def __getitem__(self, item):
>         if type(item) == types.SliceType:
>             if item.start is None:
>                 start = 0
>             else:
>                 start = item.start
>             if item.step is None:
>                 step = 1
>             else:
>                 step = item.step
>             if item.stop is None:
>                 stop = sys.maxint
>             else:
>                 stop = item.stop
>             return xrange(start, stop, step)
>         elif type(item) == types.IntType:
>             return item
>         else:
>             raise ValueError, "XRanges needs to be sliced with indices or slices"

Actually, they can be sliced with tuples too (try Int[1:2, 4:5]). But you
get the idea. I don't think a change in Python syntax is needed to clean
up range and xrange.

Allowing slice indices to be any objects, and not just integers...
that could be a nice addition. Like in the Enumerator metaclass example:
you could have

    for x in Color[red:blue]: ...

>     def __getslice__(self, start, stop):
>         return self.__getitem__(slice(start, stop, None))
> 
> Int = XRangeClass()



From dhtold at wp.pl  Tue Nov 16 14:57:33 1999
From: dhtold at wp.pl (Damian Haustein)
Date: Tue, 16 Nov 1999 20:57:33 +0100
Subject: How to close socket ?
Message-ID: <3831B72D.2447B7FC@wp.pl>

Hi,

I've a problem. I don't know how to close socket on forked process when
i kill parent process.

greets DHTold :)



From gmcm at hypernet.com  Wed Nov 24 08:20:15 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 24 Nov 1999 08:20:15 -0500
Subject: Python complaints
In-Reply-To: <383BC9C2.901A2D6@python.net>
Message-ID: <1268672881-3462615@hypernet.com>

Thomas A. Bryan wrote:

> Mikael Olofsson wrote:
> > 
> > On 24-Nov-99 Thomas Hamelryck wrote:
> >  >  Block delimitation by indenting. 
> >
> > I am one of those who appreciates this. My reason is the
> > following:
> 
> I generally don't mind this aspect of Python, but I've recently
> found that it can cause problems when transmitting code.  For
> example, I saw some code that I wanted to copy in DejaNews c.l.py
> archives, but the lines wrap in narrow columns in DejaNews.  I
> had to reindent the code.  Grrr.
> 
> I have also had problems when e-mailing code when I e-mail the
> code as text in the message instead of attaching a document.  
> 
> Is there a better way to transmit code so that possible
> indentation munging won't force the reader to reindent? 

You're thinking of Tools/scripts/pindent.py.

The major no-no for email is using tabs in the indentation, 
since most email clients happily turn them into spaces. See
http://www.python.org/ftp/python/contrib/System/tabcleaner.py
 (among others) for a tool to transform indentation. Doesn't 
deal with wrapped lines, but that's a comparatively minor 
irritation.

- Gordon



From geoff at gnaa.net  Sat Nov 13 17:00:02 1999
From: geoff at gnaa.net (geoff nordli)
Date: Sat, 13 Nov 1999 14:00:02 -0800
Subject: NT admin scripts with python
Message-ID: <000101bf2e22$7060a190$020a10ac@p350.gnaa.net>

Does python have a module that would allow a person to write
scripts on NT Server to lets say:  create users, map drives, 
modify registry, create shares, and so on.

thanks,

Geoff Nordli




From robin at alldunn.com  Mon Nov 15 15:22:30 1999
From: robin at alldunn.com (Robin Dunn)
Date: Mon, 15 Nov 1999 12:22:30 -0800
Subject: Production Web Deployment?
References: <NBBBKMMANMGEEBHPJJECGEGIFJAA.dirkl@home.com> <3826C69F.B4E26B89@digicool.com> <14379.18122.156471.644349@ghost.bbn.com> <382D68D3.46CEFD61@digicool.com> <3d4seqp52n.fsf@amarok.cnri.reston.va.us>
Message-ID: <m6_X3.674$pw2.15597867@news.randori.com>

>
> So, does this mean that if I use the FCGIPublisher, I can have
> multiple threads in Zope at the same time, without having to run
> ZServer?  Good news, if so...
>

I don't know about FCGIPublisher, but recent work has been in making ZServer
connectable via FastCGI.


--
Robin Dunn
Software Craftsman
robin at AllDunn.com
http://AllDunn.com/robin/
http://AllDunn.com/wxPython/  Check it out!






From aaron_watters at my-deja.com  Fri Nov  5 10:30:12 1999
From: aaron_watters at my-deja.com (aaron_watters at my-deja.com)
Date: Fri, 05 Nov 1999 15:30:12 GMT
Subject: Just a note
References: <381D5E84.98931682@bibsyst.no> <381D78C3.9BBECCB0@bibsyst.no>
Message-ID: <7vut5l$pgn$1@nnrp1.deja.com>

In article <381D78C3.9BBECCB0 at bibsyst.no>,
  thomas at bibsyst.no wrote:
> Hi,
>
> I could use stuff like MySQL too, if it`s easy to install and use with
> Python, but since I only want to store objects and fetch them by a
single
> key, the simplest way to do this would be using some of the built-in
stuff
> in Python.

Not a bad idea, give it a shot.

>
> I will store HUGE amounts of data, so I think Gadfly would be "too
small"
> as DBMS, since it stores all it`s data in memory ( as I recall ) .
>
> PS! If this mail made no sense, try reading the message prior to this
one.

As mentioned in another post, if you can divide the data into the
unstructured "big part" and a structured small part, gadfly might
still be useful.  Also, please have a look at bplustree.py which
provides many of the advantages of berkeley db but in a pure python
implementation.

  http://starship.skyport.net/crew/aaron_watters/bplustree/

  -- Aaron Watters http://www.chordate.com

===

"Dear Winston, enclosed find two tickets to the first
performance of my latest play.  Come and bring a friend,
if you have one.  -- G.B. Shaw"

"Dear George, I regret that I cannot attend the first
performance.  I will attend the second,
if there is one.  -- W. Churchill"


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



From tre17 at student.canterbury.ac.nz  Thu Nov  4 07:36:04 1999
From: tre17 at student.canterbury.ac.nz (Tim Evans)
Date: 05 Nov 1999 01:36:04 +1300
Subject: Python and Literate Programming
References: <Pine.LNX.3.96.991104011425.943A-100000@dhcp9545112>
Message-ID: <c75iu3il9hn.fsf@pc142.cosc.canterbury.ac.nz>

"Philip H. Jensen" <pjensen at columbus.rr.com> writes:

> Greetings.  I'm a recent convert to Python, and wish I had looked into
> it years ago.  As I said in a posting on my local Linux UG's mailing
> list, "I like everything about Python, whereas there are many things
> about Perl that I don't like."
> 
> Incidentally, I'm also a booster for Modula-3, and of course Python
> has well-known and intentional similarities to M3.  But one of my other
> enthusiasms is Literate Programming, and I'm wondering about the
> possibilities for it in Python.
> 
> Documentation texts can be easily accomodated, as can pretty-printing.

Cross-referencing on the other hand you will probably have to abandon
(at least for most stuff).  Dynamic object oriented languages make it
practically impossible.

> But for me, the core of LP, which I use (e.g. in CWEB) even if I never
> run the program through CWEAVE, is >> factoring <<.
> 
> Here's how it might look in a Python-friendly form.
> 
> 	if not dictionary.has_key(word):
> 	    <Inform the user that |word| is not present in
> 		the dictionary, and take action as directed>
> 
> then elsewhere in the program:
> 
> 	<Inform the user that |word| is not present...>:
> 	    code...
> 

For just this functionality, you could try using the `noweb' tool.  It 
manages to not break pythons indentation much at all.  For example,
the input:

<<example outer block>>=
for p in parrots:
    <<complain if the parrot p is dead>>

<<check if the parrot p is alive>>=
if p.perched and not p.nailed:
    print "It's f****** stuffed it!"

Will actually be correctly turned into:

for p in parrots:
    if p.perched and not p.nailed:
        print "It's f****** stuffed it!"

I'm not sure if other tools will do this correctly.  Noweb has the
disadvantages that it's written in icon, and that its LaTeX output is
not great, but it isn't all bad.

> ***
> Now, personally I would like to see this in the language itself,
> but that seems unlikely.  What would be essential, though would be
> 
>   1) a hook whereby "import foo" would see if there was a "foo.pyw"
>      more recent than "foo.py", and if so, would invoke a pytangle
>      program, and
>   2) a means for file and line information to be passed along, so
>      that error messages would refer to the .pyw source.
> 
> Any interest?  If I myself were to work on patches along these
> lines, could they become candidates for inclusion in the mainstream
> source tree?
> 
>     (I have some other Python development ideas,
>      but will save them for later postings.)
>   Phil Jensen
>   pjensen at columbus.rr.com 

--
<half-way-through-an-all-nighter-to-finish-a-literate-
 -programming-cs-honours-projectly>-yours
Tim Evans



From m.faassen at vet.uu.nl  Sun Nov 14 06:06:11 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 14 Nov 1999 11:06:11 GMT
Subject: A Modest Proposal -- <wink>
References: <99110613580407.02970@quadra.teleo.net> <8072o6$m4i$1@newshost.accu.uu.nl> <y0jiu3b2357.fsf@vier.idi.ntnu.no> <80jqqo$n3m$1@newshost.accu.uu.nl> <y0j9042tlrx.fsf@vier.idi.ntnu.no>
Message-ID: <80m533$mvm$1@newshost.accu.uu.nl>

Magnus L. Hetland <mlh at vier.idi.ntnu.no> wrote:
> m.faassen at vet.uu.nl (Martijn Faassen) writes:

[snip]
>> It defines a XPPMS (eXtensible PreProcessor Macro System), which allows
>> single <wink> tags. The XPPMS system transforms the source into valid XML.
>> After that the other subsystems of XWTSNTSDL take over.

> But - if you have XPPMS, why do you need XWTSNTSDL? You could just
> transform

>    This is some text <wink>

> into

>    <wink>This is some text</wink>

> and Bob's your uncle... Then you can use standard XML tools... :)

> Or - does XWTSNTSDL include tools for deciding where the wink-section
> starts, semantically?

Yes, you got it. :) It also takes care of wink levels <.50 wink>. 
The XWink subset of XWTSNTSDL takes care of any implied hypertext references.

Anyway, of course I'm spoofing the plethora of XML (draft) standards that I
can't yet make much of (I'm up to XML and the basic XML dom now :).

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?



From fredrik at pythonware.com  Tue Nov 30 10:38:52 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 30 Nov 1999 16:38:52 +0100
Subject: bug in tkinter canvas event handling?
References: <38429BB0.1BFFB54B@sage.att.com> <3843C0A0.94990E64@home.com> <3843E3D9.9C68E2A1@callware.com>
Message-ID: <010c01bf3b49$0225b420$f29b12c2@secret.pythonware.com>

Ivan Van Laningham <ivanlan at callware.com> wrote:
> > > seems simple, but i don't get what i expect.  in particular,
> > > if i press a mouse button while pointing to one object, then drag
> > > to another and release, the release event callback tells me i
> > > was pointing to the first object.  i've distilled it down to a
> > > small test case, attached below.
> > >
> > > this seems like a bug to me, but experience has shown that it's
> > > probably a misconception on my part about how the event handling
> > > is supposed to work.
> > 
> > This behavior surprised me at first, too.  It's as though the first
> > canvas object is grabbing the pointer so that all events go to it.  I
> > got around the problem by looking for the x,y coords of the release
> > event and then finding the object(s) under that point on the canvas.
> 
> Actually, this isn't a bug.  All windowing systems (the ones I know
> about, anyway) behave this way by design; the down-click does indeed
> grab the mouse pointer, and does not let go until the up-click (or is
> that up-chuck?).

yep, but that's usually limited to entire windows -- that is, if you
press a mouse button over a window, all events will be sent to
that window until you release the mouse button.  Tk's canvas
widget contains special code to handle this for items inside a
single window (see PickCurrentItem in tkCanvas.c for details).

> You can see this behaviour even in windows, which
> hacks away at expected pointer motion with gleeful abandon.
> Click on a scrollbar, begin moving it, move the pointer out of
> the scrollbar, *watch the scrollbar snap to its previous position*

note that under Windows, it only snaps back if you
move it far enough.

> move the pointer back into the scrollbar, and finally *watch the
> scrollbar snap back under the pointer.*  This is repulsive and in-
> correct behaviour

or in other words, it's there by design: it allows you to
quickly get back to the original position, if you should
change your mind.  it also makes it easy to jump back
and forth between two positions in a long list

doesn't really have anything to do with the automatic
mouse grab -- the widget simply ignores motion events
if the mouse is too far away from the scrollbar (also
note that some applications, like MSIE, doesn't use this
feature.  it's most useful if the scrolled view can be very
large, and if most items in that view look about the same).

</F>




From scottrharris at earthlink.net  Fri Nov 12 02:33:35 1999
From: scottrharris at earthlink.net (scott harris)
Date: Fri, 12 Nov 1999 03:33:35 -0400
Subject: QuickDraw 3D and Python?
Message-ID: <121119990333352139%scottrharris@earthlink.net>

Has anyone written a python wrapper around QuickDraw 3D?  
I need to generate and display 3D models from within a python
program.  I've though about writing C code to make a 3DMF file and
using applescript to launch a separate 3D viewer to give the illusion
of
integration.

Is there a simple 3D graphics sytem that works with Python and the mac?

Please respond to me on here and via email if you can be of any help.

Thanks

-- 
-Scott Harris
scottrharris at earthlink.net




From darcy at vex.net  Thu Nov 11 12:05:16 1999
From: darcy at vex.net (D'Arcy J.M. Cain)
Date: 11 Nov 1999 17:05:16 GMT
Subject: ASCII delimited files
References: <942231502.1914033961@news.interbusiness.it> <80ekhb$1qrh$1@hub.org>
Message-ID: <80et0b$2df8$1@hub.org>

D'Arcy J.M. Cain <darcy at vex.net> wrote:
> I couldn't find this on the Python or Starship sites but you can get it
> from http://www.deja.com/getdoc.xp?AN=448769706.  One small problem with
> it is that it doesn't handle commas within quoted strings properly.  I
> am going to look at this today for another project and I will post it
> here when I have something.

And here it is.  Note that Christian Tismer wrote the original (at least
he posted it) but didn't leave his name in the script itself.  I didn't
want to add it in since he may have had his reasons but I wanted to
acknowledge that he wrote it.

One thing I didn't change was the replacement of newlines with commas
although I'm not sure that that is the best thing to replace it with.
In fact, why replace it at all?  I may change this for myself.

Oh, if anyone has a suggestion for a more efficient way to handle
embedded commas I sure would be interested.


#! /usr/bin/env python
# delimited.py
# Modified from the original by D'Arcy J.M. Cain <darcy at druid.net>
#
# CT990226 V0.1

"""
breaking tab delimited or CSV text into a list of tuples.
Original version: SMC databook plugin, readxls.py
Generalized to be usable for any delimiter but \n.
"""

import string

def split_delimited(s, delimiter=None) :
	"""split a delimited text file string into a list of tuples.
	Quotes are obeyed, enclosed newlines are expanded to tab,
	double quotes go to quotes"""

	# 980426 finding line delimiter dynamically

	probe = s[:10000]
	eol = findlinedelimiter(probe)

	# 990226 guessing field delimiter from '\t,' if not supplied

	if not delimiter:
		candidates = [
			(string.count(probe, '\t'), '\t'),
			(string.count(probe, ','), ','),
		]
		candidates.sort()
		delimiter = candidates[-1][-1]
	del probe

	# the trick to make this handy is to use \0 as a placeholder

	# Kind of ugly but it works for embedded commas - DJMC
	inquote = 0
	for i in range(len(s)):
		if s[i] == '"': inquote = (inquote == 0)
		if inquote == 0 and s[i] == delimiter:
			s = s[:i] + "\0" + s[i + 1:]
	parts = string.split(s, '"')
	limits = (0, len(parts)-1)
	for i in range(len(parts)) :
		# may as well strip the spaces now - DJMC
		part = string.strip(parts[i])
		if i%2 :
			part = string.replace(part, eol, delimiter)
		else :
			if not part and i not in limits: part = '"'
		parts[i] = part
	# merge it back
	txt = string.join(parts, "")
	parts = string.split(txt, eol)
	# now break by \0
	for i in range(len(parts)) :
		fields = string.split(parts[i], "\0")
		parts[i] = tuple(fields)
	return parts

# utilities

def findlinedelimiter(txt) :
	"""
	provide some kb of text to this function. It will determine
	the best delimiter and therefore guess the system
	"""
	mac  = "\x0D"
	unix = "\x0A"
	dos  = mac+unix
	oses = [dos, unix, mac]
	# find the one which gives the most lines.
	# in doubt, the longest delimiter wins.
	lis = []
	while txt and txt[-1] in dos: txt = txt[:-1] # CT970904
	for delim in oses:
		lis.append((len(string.split(txt, delim)), delim))
	lis.sort();
	return lis[-1][-1]

if __name__=="__main__":
	for l in  split_delimited('''1, 2,3, "vier", "quo""te", "embedded, comma", "this
is with a newline", here
another record'''):
		print l

# eof



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



From news at dorb.com  Wed Nov 24 09:57:46 1999
From: news at dorb.com (Darrell)
Date: Wed, 24 Nov 1999 09:57:46 -0500
Subject: Python complaints
References: <XFMail.991124113221.mikael@isy.liu.se> <383BC9C2.901A2D6@python.net>
Message-ID: <Y%S_3.2756$X5.263232@typhoon.nyroc.rr.com>

Use spaces instead of tab's.
I have a script to convert tabs to spaces.
This fixed most of my concerns with white space block structure.

--
--Darrell

"Thomas A. Bryan" <tbryan at python.net> wrote in message
news:383BC9C2.901A2D6 at python.net...
> Mikael Olofsson wrote:
> >
> > On 24-Nov-99 Thomas Hamelryck wrote:
> >  >  Block delimitation by indenting.
> >
> > I am one of those who appreciates this. My reason is the following:
>
> I generally don't mind this aspect of Python, but I've recently found that
> it can cause problems when transmitting code.  For example, I saw some
> code that I wanted to copy in DejaNews c.l.py archives, but the lines
> wrap in narrow columns in DejaNews.  I had to reindent the code.  Grrr.
>
> I have also had problems when e-mailing code when I e-mail the code as
> text in the message instead of attaching a document.
>
> Is there a better way to transmit code so that possible indentation
> munging won't force the reader to reindent?
>
> vaguely-remembering-something-about-option-BEGIN-and-END-ly yours
> ---Tom
>
> --
> http://www.python.org/mailman/listinfo/python-list





From moonseeker at my-deja.com  Tue Nov 16 03:46:34 1999
From: moonseeker at my-deja.com (moonseeker at my-deja.com)
Date: Tue, 16 Nov 1999 08:46:34 GMT
Subject: Daemonic threads
References: <3830C73F.C8A7CA62@DONT.LIKE.SPAM>
Message-ID: <80r5la$ikv$1@nnrp1.deja.com>

In article <3830C73F.C8A7CA62 at DONT.LIKE.SPAM>,
  "David S." <I at DONT.LIKE.SPAM> wrote:
>
> I'm writing a CGI script that will launch a long-running process in the
> background, and return some HTML page to the browser.  Now, I know how
> to do that with 'fork()', but it seems to me that I should be able to do
> the same thing with a daemonic thread from the 'threading' module.
>
> I can't get that to work, though,
>
>    ...
>    xt = threading.Thread(target=_threadExec, args=(save, mail, None))
>    xt.setDaemon(1)
>    xt.start()
>    ...
>
> and the script exits with no thread running.  Non-daemonic threads seem
> to work as I expect them to.  What don't I understand about daemonic
> threads?  (Python 1.5.2 on Linux 2.2.7, if its relevant.)

Hi!

Daemonic threads die if no other non-daemonic thread lives. At the moment
your script (the main thread, non-daemonic) dies, all (daemonic) child
threads die too.

Mike


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



From jon at rdt.monash.edu.au  Wed Nov  3 21:52:50 1999
From: jon at rdt.monash.edu.au (Jonathan Giddy)
Date: 4 Nov 1999 02:52:50 GMT
Subject: Exceptions from callbacks
References: <381CA8E5.5D69E64@lords.com> <7virj4$b7b$1@towncrier.cc.monash.edu.au> <381FAE6D.73F426FE@lords.com> <19991103071727.A87523@vislab.epa.gov> <3820DE5B.A5500330@lords.com>
Message-ID: <7vqse2$4km$1@towncrier.cc.monash.edu.au>

Eric Dorland <dorland at lords.com> writes:

>I understand what you mean exactly. The problem is I can't do it that
>way. Basically, the way the library works is that you register a bunch
>of callbacks and then call a function (eg. Loop()) that doesn't return.
>The HandlePendingEvents (its equivalent I mean :)) isn't exposed by the
>library, so I can't wrap it :). That's way I want a way to force the
>interpreter to realize an exception is there. Is there a way, or a work
>around, to do this?

Are you saying that there is no way to tell the function "Loop" to return?
Forgetting Python for a moment, can you even write a C program that breaks 
out of the loop to handle errors?

If so, then you can change Randall's example callback to the code below.

If not, then you have the problem that the Python interpreter *is* a C 
program.  If a C program can't do it, then neither can Python.

  static void _wrap_MYLIBCallback( void )
    /*  C callback used to relay to Python callable objects.  */
  {
    ...

    /*  Pass the buck onto Python  */
    pyresult = PyEval_CallObject( pyfunc, pyargs );     /*  Call Python     */
    Py_DECREF( pyargs );                                /*  Trash arg list  */
    if ( pyresult )
      Py_DECREF( pyresult );                            /*  Trash any result*/
    else
      /* An exception occurred, do whatever tells the loop to exit */
      QuitLoop();

    /*  If the callback threw an exception, it will be propagated across    */
    /*    the C layer ( we check PyErr_Occurred() on the other side).       */
  }




From smst at quantisci.co.uk  Fri Nov 26 06:53:27 1999
From: smst at quantisci.co.uk (Steve M S Tregidgo)
Date: Fri, 26 Nov 1999 11:53:27 GMT
Subject: dir() and propertly list NEWBIE
In-Reply-To: <81f54p$ig4$1@nntp6.u.washington.edu>
Message-ID: <199911261153.LAA07241@cobweb.quantisci.co.uk>

Donn Cave wrote:
> Quoth tiddlerdeja at my-deja.com:
> [snip]
> | Similarly, If I can get an objects functions, is it possible then
get a
> | list of the possible parameters for that function?
> 
> No!  Wouldn't that be nice?
> 

Take a look at the function's attribute 'func_code'.  Within this
you'll find the information you want.  func_code's attributes include
'co_varnames' and 'co_argcount'.  The first co_argcount items in
co_varnames give you the parameters.

>>> import rfc822
>>> rfc822.parseaddr.func_code.co_varnames
('address', 'a', 'list')
>>> rfc822.parseaddr.func_code.co_argcount
1
>>>

So the parameters to rfc822.parseaddr are ('address', 'a',
'list')[:1]
-- or just 'address'.

(Note:  func_code.flags also tells you something here.  If flags&4 is
true then the function takes a *arg type parameter:

 def f(a, b, *c):

The name of the relevant parameter is the next unused item in
co_varnames.  If flags&8 is true then it also takes keyword
arguments:

 def f(a, b, *c, **kw):

Again, the next unused item in co_varnames is the parameter.)

Default arguments are defined in the bytecode of the function's
parent namespace -- they're harder to get hold of...

Have fun,
Steve Tregidgo



From edwardam at home.com  Tue Nov 30 23:09:45 1999
From: edwardam at home.com (Edward Muller)
Date: Wed, 01 Dec 1999 04:09:45 GMT
Subject: [Zope-dev] Re: Exposing COM via XML-RPC or Something Else
References: <613145F79272D211914B0020AFF64019276318@gandalf.digicool.com>
Message-ID: <38449F86.D6538558@home.com>

Michel Pelletier wrote:

> > -----Original Message-----
> > From: Thomas Heller [mailto:thomas.heller at ion-tof.com]
> > Sent: Tuesday, November 30, 1999 2:52 PM
> > To: Brian Lloyd; 'Edward Muller'; python-list at python.org
> > Cc: zope-dev at zope.org
> > Subject: [Zope-dev] Re: Exposing COM via XML-RPC or Something Else
> >
> >
> > > You might want to take a look at MS' SOAP (Simple Object Access
> > > Protocol) spec - it's not fully baked yet, but it basically takes
> > > xml-rpc several steps further, allowing you to communicate with
> > > SOAP-enabled objects (COM or otherwise) using XML over HTTP. We in
> > > Zope-land are keeping a close eye on this, as publishing objects
> > > via HTTP is the core idea of Zope already.
> >
> > Zope and SOAP sounds very similar.
> > Is this by accident?
>
> Yup
>
> -Michel

I took a look at the SOAP specs at work, and it is basically XML-RPC with
some differences....It looks like MS is at it again....designing YAAPI
(Yet Another API) to confuse the world....

:-(






From usenet-mm-yyyy at julesallen.com  Fri Nov 19 07:19:30 1999
From: usenet-mm-yyyy at julesallen.com (Jules Allen)
Date: Fri, 19 Nov 1999 12:19:30 GMT
Subject: The First Annual Obfuscated Python Contest
Message-ID: <mfbZ3.71$AZ6.1807@dfiatx1-snr1.gtei.net>

I'm kidding, of course.

But while doing my morning grep of Slashdot, I did come across this:

	http://slashdot.org/comments.pl?sid=99/11/17/2148249&
	threshold=0&commentsort=0&mode=thread&pid=12#36

(rejoin that line)

Eeek!

Who-said-python-was-always-clean-ly-yrs,
Jules

-- 
Want to reply via email? Put the current month and year in 
my email address.



From greg.ewing at compaq.com  Thu Nov  4 04:24:48 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Thu, 04 Nov 1999 22:24:48 +1300
Subject: What are Aspects? (Re: How can I make my assertions smarter?)
References: <381D7087.BF8510E5@webone.com.au> <slrn8216ve.j7h.wtanksle@hawking.armored.net>
Message-ID: <382150E0.F8938A7A@compaq.com>

William Tanksley wrote:
> 
> Aspects are really cool -- like a semantic preprocessor on steroids.

Where can we find out more about these Aspect thingies?

Greg



From skaller at maxtal.com.au  Fri Nov 12 05:31:21 1999
From: skaller at maxtal.com.au (skaller)
Date: Fri, 12 Nov 1999 21:31:21 +1100
Subject: __str__ vs. __repr__
References: <000001bf275c$02370d20$1d2d153f@tim> <7vvni7$19q0$1@nntp6.u.washington.edu>
Message-ID: <382BEC79.EC65913D@maxtal.com.au>

 > Quoth "Tim Peters" <tim_one at email.msn.com>:
> 
> | Let me back off to what repr and str "should do":
> |
> | repr(obj) should return a string such that
> |     eval(repr(obj)) == obj

FYI: Viperi (my python interpreter) contains an extension, a function
erepr,
which provides these semantics for a much wider
class of types than CPython 1.5.2, including: modules, classes,
instances,
and functions, provided all the attributes are themselves
'erepr'able. Erepr will not handle (native) file objects.

There is, however, a caveat. 

	erepr(f)

returns a string representing the definition of the function f.
However, evaluating the string will NOT work correctly,
unless the globals is correctly set, something like:

	eval(erepr(f), f.__module__)

is required. Similarly for classes.

Oh yes, I forgot to mention that all Viper statements are also
expressions, the 'e' in erepr means 'expression'.
In fact, given an expression, such as 

	x + 1

then, erepr applied to this expression yields a string

	'x + 1'

There is currently no way to define such an expression in Viperi.
[A 'lazy' or 'defered' evaluation control is under consideration]
However, it has facilities for partial evaluation, they're just
disabled for Python compatibility. For example, with partial
evaluation enabled, the result of:

	y = 1
	z = x + y

is that x is an _expression_ whose value is

	x + 1

In Viper, expressions are THE first class type. All other types
are cases of expressions (including statements, objects, and values
of almost all kinds -- slices are not expressions at the moment).

You may be wondering how this works. The answer is that Viperi does not
use bytecode, it executes the 'Abstract Syntax Tree' directly.
Consequently, the original code can be recovered. (without comments,
or the original formatting, of course).

Note that this will NOT work if the object has been bound,
that is, if names have been replaced by numeric indexes.
(And, it will not work in compiled code either, for the same reason:
bound or compiled code is the 'moral equivalent' of python bytecode)

| I'd go further and formalize a property that's currently honored but
> | implicitly:  the string returned by repr should contain only characters c
> | such that

Viper repr functions uses only 'printable characters' in the range
33-126
of ISO10646, plus space. (i.e. printable ASCII characters plus space).
All other codes are represented either using \n (and related specials)
or using \uXXXX or \uXXXXXXXX encoding. [There will be a 
python compatibility switch to force use of python lexicology]

-- 
John Skaller, mailto:skaller at maxtal.com.au
1/10 Toxteth Rd Glebe NSW 2037 Australia
homepage: http://www.maxtal.com.au/~skaller
downloads: http://www.triode.net.au/~skaller



From jom at scf.usc.edu  Tue Nov 16 19:36:51 1999
From: jom at scf.usc.edu (juwon om)
Date: Wed, 17 Nov 1999 00:36:51 -0000
Subject: Why my Pathyon GUI is not woking?
Message-ID: <80tphr$jta$1@usc.edu>

Python GUI( for win98 NT)is not working suddenly. Anyone have idea?

It was working well.







From bernhard at alpha1.csd.uwm.edu  Mon Nov 29 13:49:45 1999
From: bernhard at alpha1.csd.uwm.edu (Bernhard Reiter)
Date: 29 Nov 1999 18:49:45 GMT
Subject: XPython (X11)
References: <383E5350.1E07536A@xrce.xerox.com> <slrn83tj53.k7h.bernhard@alpha1.csd.uwm.edu> <81ms2n$126$1@essle.valdyas.org> <slrn840dt7.mov.bernhard@alpha1.csd.uwm.edu> <m3hfi64oyh.fsf@phaedrus.zen.xom>
Message-ID: <slrn845im9.178.bernhard@alpha1.csd.uwm.edu>

On 29 Nov 1999 01:26:30 +1000, Rob Hodges <s323140 at student.uq.edu.au> wrote:
>bernhard at alpha1.csd.uwm.edu (Bernhard Reiter) writes:
>
>> >Hey! Don't forget PyQt and PyKDE (and PyQt seems to be
>> >working under Windows, too, nowadays)...
>> Do you have to buy the Troll Kit first? >:)
>> 
>> And I didn't mention pyGNOME, why should I mention to PyKDE?
>> To create an application tied to KDE or GNOME is a mistake in my
>> eyes anyway. <wink>

>I'm struggling a bit to understand why you say that in this context.
>I can't speak for pyKDE (although I imagine the situation is the
>same), but developing with pyGnome doesn't imply a 'tying' to Gnome;
>only that the Gnome libs must be available in addition to the GTK libs
>-- just as you must have wxwin in addition to GTK to run wxpython
>apps.  One advantage of PyGnome over straight PyGtk is that it offers
>the Gnome Canvas widget, which can be useful whether the user is
>running Gnome or not.

>There is also no harm in taking advantage of the various services
>provided by these desktop environments to improve the user's
>experience, provided that your app also works nicely without them;

True. But I want that over an integrated API and GNOME and KDE
don't offer that yet. And they also don't offer it cross platform.
wxWindows at least tries to get there.

>And yeah, last I heard you had to buy the Troll Kit for PyQt under
>Win32.  
Dropping platform compatibility, that a huge point.

>I strongly prefer PyGtk over WxPython on Linux, so I'm
>curious in case I should have to develop anything for Win in the
>future (but, I guess, not curious enough to reboot :)

Hmm wxPython seems very reasonable to me, why not fix that then.
	Bernhard
-- 
Research Assistant, Geog Dept UM-Milwaukee, USA.  (www.uwm.edu/~bernhard)
Free Software Projects and Consulting 		         (intevation.net)  
Association for a Free Informational Infrastructure            (ffii.org)



From nascheme at enme.ucalgary.ca  Tue Nov 16 19:03:57 1999
From: nascheme at enme.ucalgary.ca (Neil Schemenauer)
Date: Wed, 17 Nov 1999 00:03:57 GMT
Subject: ANNOUNCE: a few little things
Message-ID: <slrn833s87.elm.nascheme@adler.dynodns.net>

I spent a bit of time today and updated my Python web site.  The
highlights are:


py152_sa.zip
------------

A Stand-alone version of Python for Windows. It includes Python
1.5.2, Tcl, zlib and bsddb.  It fits on a floppy and can be
extracted and run from any location.  Just add your application
code. 

It is very possible that I screwed something up when making the
zip file.  If you have problems running the "hello world"
application on a machine without Python installed, please let me
know.


sparsemodule-0.3.tar.gz
-----------------------

An extension module wrapping the sparse library.  It can be used
for solving large systems of linear equations.  Included is a
compiled Windows binary.  Changes in this release are:

    0.2 to 0.3
    ----------

       - fixed off by one bug in set_element
       - fixed bugs in spDestroy and RecordAllocation
       - added test2.py program from D. Michael McFarland 
         <mcfarlan at caber-eng.com>
       - added Matrix.py and LUP.py
       - moved sparse routines into their own directory


py-m4.tar.gz
------------

m4 macros for creating extension classes.  This is very alpha.  I
don't expect anyone to actually use them.  I just want to know if
anyone is interested.  They definitely make simple extension
modules easier to write.

Here is some sample code for creating a Hello module with a Hello
class with a hello method:

= hello.m4 ==========================================================
m4_include(py.m4)

PY_MODULE(Hello)

PY_CLASS(Hello, HW)
        PY_DEF(hello, self, args)
        {
                Py_INCREF(Py_None);
                return Py_None;
        }

PY_END_CLASS
=====================================================================


This software can be downloaded from:

    http://www.ucalgary.ca/~nascheme/python/

If you have any problems please let me know.


    Neil

-- 
"The percentage of users running Windows NT Workstation 4.0 whose PCs stopped
working more than once a month was less than half that of Windows 95 users."
  -- microsoft.com/ntworkstation/overview/Reliability/Highest.asp



From kaloian at europe.com  Fri Nov 26 05:39:53 1999
From: kaloian at europe.com (Kaloian Doganov)
Date: Fri, 26 Nov 1999 11:39:53 +0100
Subject: Python extensions and lcc-win32
Message-ID: <383E6378.4B10FDD6@europe.com>

Can I use lcc-win32 to write Python extensions? The .h files in include/
directory provided with Python 1.5.2 (Win32) does not seems to be
compatible with lcc-win32.

I'm using lcc-win32 v3.1 (Nov 17 1999)

In fact, I'm looking for a non-MSVC solution to write Python extensions.
So, any GCC or lcc-win32 tips are appreciated.






From garry at research.att.com  Fri Nov  5 10:50:52 1999
From: garry at research.att.com (Garrett G. Hodgson)
Date: Fri, 5 Nov 1999 15:50:52 GMT
Subject: tkinter grid() question
References: <381F2D32.9E3A15CF@research.att.com> <381FFF52.707400E5@compaq.com>
Message-ID: <3822FCDC.F5C5462D@research.att.com>

Greg Ewing wrote:
> 
> "Garrett G. Hodgson" wrote:
> >
> > in the code below, i create a Text entry area with a Scrollbar.
> > i want this pair to always use all of the available space.
> > but when i resize the window, the stay the same size, neither
> > expanding or shrinking.
> 
> You have to do two things: make the Text widget sticky on
> all sides, and configure the weight of the row and column
> containing the Text widget so that it sucks up all the
> extra space.
> 
> It's simplest if the widgets only take up one cell of
> the grid each, so I'll show you how to do it that way:
> 
>    self.body.grid(row = 0, column = 0, sticky = N+S+E+W)
>    scrollbar.grid(row = 0, column = 1, sticky = N+S)
>    self.rowconfigure(0, weight = 1)
>    self.columnconfigure(0, weight = 1)

bingo.  that did it.

> If you're going to put other widgets in the frame as
> well, you'll have to be careful how you configure the
> grid so that resizing the text widget properly doesn't
> have undesirable side effects on the layout of the other
> widgets.

indeed, my orignal code had more stuff in it. i'd simplified
it for purposes of the question.
thanks very much for your help.


-- 
Garry Hodgson			"Hey, mister, can ya tell me,
garry at sage.att.com		where a man might find a bed?"
Software Innovation Services	He just grinned and shook my hand,
AT&T Labs			"No", was all he said.



From mikael at isy.liu.se  Thu Nov 25 04:37:13 1999
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Thu, 25 Nov 1999 10:37:13 +0100 (MET)
Subject: Upgrading Python on Windows
In-Reply-To: <iQR_3.5212$VA6.25796@news-server.bigpond.net.au>
Message-ID: <XFMail.991125103713.mikael@isy.liu.se>

On 24-Nov-99 Mark Hammond wrote:
 > >I-can-troubleshoot-a-UNIX-install,-but-Windows-scares-me-ly yours
 >  
 >  You should be able to uninstall Python 1.5.1 from Control Panel->Add Remove
 >  Programs.  Id suggest doing before installing the new version.
 >  
 >  Nothing-to-be-scared-of-ly,

Make a backup and/or check your disk recovery toolbox first.

I recall the first time I tried to uninstall a program under Win95. My
entire Programs directory disappeared into some black hole.

Certainly-something-to-be-scared-of,-it's-MS-ly y'rs

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    25-Nov-99
Time:    10:30:36

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



From c.evans at clear.net.nz  Wed Nov 17 02:32:50 1999
From: c.evans at clear.net.nz (Carey Evans)
Date: 17 Nov 1999 20:32:50 +1300
Subject: Bug? - win32pipe.popen2 is moody under W98
References: <80qkbv$2m5$1@nntp5.atl.mindspring.net> <cg6Y3.5073$Xo.26549@news-server.bigpond.net.au>
Message-ID: <871z9plgil.fsf@psyche.evansnet>

"Mark Hammond" <mhammond at skippinet.com.au> writes:

> However, it will be pretty resource hungy - the solution (as
> documented by MS) is to execute a "shim" exe - so each popen on
> Win95/98 will now have _3_ new processes - this shim, command.com
> and the actual .exe you are trying to run.

Is there any chance of a flag, or different functions, to disable the
use of command.com?  When I've tried it, the default behaviour of
CreateProcess() has done everything I've wanted, with the bonus that I
get the return code of the program.

In fact, if I knew I was calling a 32-bit app I should be able to do
without the shim, but it's also not that hard to roll my own in that
case.

-- 
	 Carey Evans  http://home.clear.net.nz/pages/c.evans/

	       "This is where your sanity gives in..."



From lincwils at teleport.com  Sun Nov 14 19:49:04 1999
From: lincwils at teleport.com (Dennis Wilson)
Date: Sun, 14 Nov 1999 16:49:04 -0800
Subject: [HELP] TKinter could not run in Python!
References: <80i63e$7t4$1@news.seed.net.tw> <memo.19991113182523.24605A@forestfield.cix.co.uk> <80mfel$p3k$1@news.seed.net.tw>
Message-ID: <aKIX3.35703$C7.1588262@news1.teleport.com>

Mark, would you be willing to post the modifications that you made under
item 3.
I found that I could get Tkinter to work by copying the tcl80.dll and
tk80.dll into my python directory, but I was then forced to re-install
tcl8.2

Thanks

Mark Wu wrote in message <80mfel$p3k$1 at news.seed.net.tw>...
>Dear David:
>    I think I already solve this problem. Because I have to use TCL/TK to
>develop software, therefore I have to keep the TCL/TK in the update to date
>version.
>    I trying to uninstall the TCL/TK v8.1 , then reinstall the TCL/TK
v8.05,
>then I make some modification of my env. and TCL source
>
>1) I set TCL_LIBRARY="the path where you install the TCL lib"
>2) then, set TK_LIBRARY="the path where you install the TK lib"
>3) Finally, I make some modification of some scripts in the 8.05 version...
>4) It works ...
>
>Thanks, I hope could share this experience with you!
>
>Best Regards, Mark
>
>






From ionel at psy.uva.nl  Tue Nov 16 11:23:28 1999
From: ionel at psy.uva.nl (Ionel Simionescu)
Date: Tue, 16 Nov 1999 17:23:28 +0100
Subject: wxPython and DISLIN work nicely together
Message-ID: <80s0b4$g5n@mail.psy.uva.nl>

Hi,


DISLIN (http://www.linmpi.mpg.de/dislin/) is a powerful visualization
package available on quite a lot of platforms, aimed at scientists.
DISLIN is "ported" to Python as well.

I want to report that wxPython and DISLIN work very well together.

Because DISLIN can paint on a window, one can easily and transparently use
it within a (X)Windows application. DISLIN can also create/convert to a
variety of formats, such as Postscript, PNG and metafiles.

Below you may find a simple illustration of the idea.
(You need both wxPython and DISLIN installed to run it.)


ionel

---

from dislin import *
from wxPython.wx import *

x = range(9)
y = x[:]

#---------------------------------------------------------------------------
class dislin_frame(wxFrame):
    def __init__(self):
        wxFrame.__init__(
            self,
            NULL, -1, "wxCML",
            wxDefaultPosition,
            wxSize(580, 400)
        )

        # paint with a double-buffering scheme
        x11mod('nostore')
        # set the DISLIN kind of paint target
        metafl('xwin')
        # set the handle of the window where DISLIN paints to us
        setxid( self.GetHandle(), 'window')
        disini()

    def OnPaint(self, event):
        plot(x,y)
        event.Skip()

#---------------------------------------------------------------------------
class myApp(wxApp):

    def OnInit(self):
        dislin_frame().Show(TRUE)
        return TRUE

#---------------------------------------------------------------------------
if __name__ == '__main__':
    app = myApp(0)
    app.MainLoop()






From wtanksle at hawking.armored.net  Thu Nov  4 15:59:23 1999
From: wtanksle at hawking.armored.net (William Tanksley)
Date: 4 Nov 1999 20:59:23 GMT
Subject: What are Aspects? (Re: How can I make my assertions smarter?)
References: <Pine.WNT.4.05.9911032042440.109-100000@david.ski.org>
Message-ID: <slrn824lqc.ngk.wtanksle@hawking.armored.net>

On Wed, 3 Nov 1999 20:43:58 -0800 (Pacific Standard Time), David Ascher wrote:
>On Thu, 4 Nov 1999, Greg Ewing wrote:
>> William Tanksley wrote:

>> > Aspects are really cool -- like a semantic preprocessor on steroids.

>> Where can we find out more about these Aspect thingies?

>www.aspectj.org

That's a good one.  The original Xerox papers (linked to from that site)
are also good.

In short, aspects are like objects, only where an object implements a
self-contained concept, an aspect implements a widespread concept.  For
example, reference counting could be written as an aspect, but never as an
object.

>--david

-- 
-William "Billy" Tanksley



From ClaraPotts at news.vex.net  Mon Nov  1 22:46:24 1999
From: ClaraPotts at news.vex.net (Clara Potts)
Date: 2 Nov 1999 03:46:24 GMT
Subject: newbie intro recs?
References: <381E3205.BEFDE12C@home.com>
Message-ID: <7vlmqg$1pi3$1@hub.org>

On Tue, 02 Nov 1999 00:32:25 GMT, Ben Wolfson <rumjuggler at home.com> wrote:
>Can someone recommend a good introduction to Python, for a very novice
>programmer?  (the only background I have is c++ from ap comp sci at high


Not sure how beginner you want to go, but is the tutorial that comes
with the standard python documentation too advanced?

    http://www.python.org/doc/current/tut/tut.html

If so here are a few other beginner ones from my small collection of
links at http://www.vex.net/~x/parnassus/apyllo.py/935043691.636055170

Learning to Program:
    http://members.xoom.com/_XOOM/alan_gauld/tutor/tutindex.htm

Non-Programmer's Tutorial:
    http://www.honors.montana.edu/~jjc/easytut/easytut/

Instant Hacking:
    http://www.idi.ntnu.no/~mlh/python/programming.html

There's also a whole whack of tutorals listed on python.org
here: http://www.python.org/doc/Hints.html

Good luck!

                





From mstenber at cc.Helsinki.FI  Wed Nov 24 08:35:48 1999
From: mstenber at cc.Helsinki.FI (Markus Stenberg)
Date: 24 Nov 1999 15:35:48 +0200
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <867lj85g4c.fsf@g.local>
Message-ID: <al8yabo9fm3.fsf@sirppi.helsinki.fi>

Gareth McCaughan <Gareth.McCaughan at pobox.com> writes:
>   - Some more iteration constructs would be nice.
> 
>       - do ... while

ugh. matter of taste I suppose, but I don't find that lack big deal.

>       - parallel iteration across more than one list

for list1item, list2item, list3item in map(None, list1, list2, list3):
	print 'whee', list1item, list2item, list3item

>       - a less yucky way of saying "for i in range(1,20,3)"

see above - although some special syntactic sugar would be "nice", I prefer
minimalistic approach over say, C.. of course, Dylan does it (IMNSHO)
cleanest:

for (i from 3 to 19 by 3)
	...
	

Regrettably I can't tolerate Dylan's whitespace approach :P
(I write accidentally too often stuff like i-j, as opposed to i - j)

> -- 
> Gareth McCaughan  Gareth.McCaughan at pobox.com
> sig under construction

-Markus

-- 
----- BEGIN GEEK CODE v3.0 BLOCK (*) -----
GCS d- s+: !a C+(++) UL+++$ P--- L+++ E++ W-- N+ W--- O---- M-- V- PS+++
Y+ PGP++>+++ t+@ 5++ X+ R* tv-- b+++ DI+++>++++ D---- G++ e++ h* r% z+
----- END GEEK CODE v3.0 BLOCK -----
* For spoilers, check http://www.geekcode.com



From endgamer71 at hotmail.com  Tue Nov  2 05:08:20 1999
From: endgamer71 at hotmail.com (Endgamer)
Date: Tue, 02 Nov 1999 10:08:20 GMT
Subject: newbie intro recs?
References: <381E3205.BEFDE12C@home.com>
Message-ID: <381eb73d.3987883@212.225.36.99>

On Tue, 02 Nov 1999 00:32:25 GMT, Ben Wolfson <rumjuggler at home.com>
wrote:

>Can someone recommend a good introduction to Python, for a very novice
>programmer?  (the only background I have is c++ from ap comp sci at high
>school, and what I've gleaned from the C++ Programming Language, so I
>can't pick things up as easily as I'd like).
>
>Thanks!
>
>-- 
>Barnabas T. Rumjuggler
>
>Graybags: "Irony = metally" (U.S. dictionary)
>Me: And here I thought it was the practice of removing wrinkles from
>clothes.
>Graybags: Removing wrinklies from clothes is euthanasia.
> -- in alt.possessive.its.has.no.apostrophe

If your going to pay for anything (dead-tree-wise), you could do a lot
worse than Mark Lutz's "Programming Python" - I've used it in
combination with some of the text files online, and I'm doing fine so
far, despite not even having the C++ experience you've got! Its
published by O'Reilley (which for me at least makes it gospel <g>) and
you can get it on amazon.com...



Endgamer
Spider Warrior

(Endgamer71 at spamtrap.hotmail.com)
#####################################
Not a quote, just an on-the-spot observation:
You know you're a geek when your PC costs exactly
15 times as much as your stereo. And gives better
sound quality.<sigh>
                              Endgamer
#####################################



From mikael at isy.liu.se  Tue Nov  2 09:01:52 1999
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Tue, 02 Nov 1999 16:01:52 +0200 (MET)
Subject: Drawing (directed) graphs
In-Reply-To: <ye1bt9das2p.fsf@ex.ac.uk>
Message-ID: <XFMail.991102160152.mikael@isy.liu.se>

On 02-Nov-99 R.M.Everson wrote:
 >  Does anyone know of Python code for drawing graphs of the
 >  graph-theoretic type? Ideally, I'd like to be able to interactively
 >  add nodes and edges etc and add data to structures associated with the
 >  nodes and edges.

I do not know of any such code, but it might very well be of interest 
to some of my colleagues.

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    02-Nov-99
Time:    15:59:03

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



From aahz at netcom.com  Mon Nov 29 18:46:02 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 29 Nov 1999 23:46:02 GMT
Subject: '==' vs. 'is' behavior
References: <382C2C6A.64A8F2C2@theriver.com> <14380.12036.197083.409565@weyr.cnri.reston.va.us> <8E7F86267duncanrcpcouk@news.rmplc.co.uk> <oqemd9dk2g.fsf@titan.progiciels-bpi.ca>
Message-ID: <81v37q$e8f$1@nntp9.atl.mindspring.net>

In article <oqemd9dk2g.fsf at titan.progiciels-bpi.ca>,
=?ISO-8859-1?Q?Fran=E7ois_Pinard?=  <pinard at iro.umontreal.ca> wrote:
>
>By the way, is there common wisdom (or rather, maybe, usage-standards-to-be)
>about using `is' instead of `=='?

"is" is faster, but you should use == unless you know for sure you want
to test object equality instead of value equality.  For example:

x = 'abc'
if x is 'abc' :
  ...

may or may not fail, depending on how python handles object creation;
OTOH, it's almost certain that the following will fail:

x = raw_input()
if x is 'abc' :
  ...
--
                      --- Aahz (@netcom.com)

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

Question authority: ask me anything!



From greg.ewing at compaq.com  Thu Nov  4 05:16:07 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Thu, 04 Nov 1999 23:16:07 +1300
Subject: A Comparison of Python and Ruby Extension Interfaces (Re: Comparison 
 between Python and "Ruby")
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp> <7vjtrc$hmm$1@newshost.accu.uu.nl> <38204614.2EAC6051@compaq.com> <021119992327527678%pereira@research.att.com>
Message-ID: <38215CE7.70C3D560@compaq.com>

Fernando Pereira wrote:
> 
> What's the effect of this on portability? For instance, how are
> pointers to live objects from the C stack handled? As is well-known,
> this cannot be done portably.

It scans the whole C stack and marks anything that looks like
it might be a pointer to a Ruby object. It assumes that setjmp()
saves any registers that might contain such pointers. There are
a few special cases coded in. Porting to a new system might
require this piece of code to be adjusted, although it looks
to me like it should work under most sensible architectures.

Global variables in C extensions which contain Ruby pointers
have to be explicitly made known during initialisation.

> > * Ruby seems to be able to raise exceptions without
> > having to explicitly propagate them back out of routines,

> Same comment. There is no portable way of unwinding the C stack. What
> does that mean for Ruby's portability?

It uses setjmp/longjmp. Presumably these work correctly on
any system where they are implemented. Porting Ruby to a
system where they're not implemented could be difficult.

Greg



From phd at phd.russ.ru  Thu Nov 25 05:33:42 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Thu, 25 Nov 1999 10:33:42 +0000 (GMT)
Subject: how to split the contain of retr (using poplib module)
In-Reply-To: <383ceefd.3147421@news.dnet.net.id>
Message-ID: <Pine.LNX.4.21.9911251030440.11528-100000@fep132.fep.ru>

On Thu, 25 Nov 1999, Ahmad Munif wrote:
> so, i can get the value of mail, such as: To,From,body message.
> how to split it ?

   Modules mimetools.py and multifile.py are your friends.

   Look http://sun.med.ru/~phd/Software/Python/misc.html, download
extract_mime.tgz and learn how this works.

   Additionals mirrors:
http://members.xoom.com/phd2/Software/Python/
http://www.fortunecity.com/skyscraper/unix/797/Software/Python/

Oleg.
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From da at ski.org  Fri Nov 26 16:53:18 1999
From: da at ski.org (David Ascher)
Date: Fri, 26 Nov 1999 13:53:18 -0800 (Pacific Standard Time)
Subject: Python complaints
In-Reply-To: <383C6A9B.4049CB12@x.x>
Message-ID: <Pine.WNT.4.04.9911261342240.198-100000@rigoletto.ski.org>

On Wed, 24 Nov 1999, Eric Jacobs wrote:

> I don't know if I like that syntax. Taking a slice
> of what? If I had to replace range, I'd go with:
> 
>    for i in Int[1:20:3]:
>       ...
> 
> Here Int is an object that represents a list of all
> integers. We take a slice of it... and get exactly
> what we expect.

> Either way, the semantics of lists with an infinite
> number of elements or negative indices would have
> to be cleared up somewhat. That would be a good
> thing.

Without dealing with the last paragraph, you can fake it now, with
something like (barely tested):

import types
class XRangeClass:
    def __getitem__(self, item):
	if type(item) == types.SliceType:
	    if item.start is None:
		start = 0
	    else:
		start = item.start
	    if item.step is None: 
		step = 1
	    else:
		step = item.step
	    if item.stop is None:
		stop = sys.maxint 
	    else:
		stop = item.stop
	    return xrange(start, stop, step)
	elif type(item) == types.IntType:
	    return item
	else:
	    raise ValueError, "XRanges needs to be sliced with indices or slices"
    def __getslice__(self, start, stop):
	return self.__getitem__(slice(start, stop, None))

Int = XRangeClass()







From parkw at better.net  Thu Nov 11 12:07:57 1999
From: parkw at better.net (William Park)
Date: Thu, 11 Nov 1999 12:07:57 -0500
Subject: .html -> .ps ?
In-Reply-To: <14378.54900.714487.334690@weyr.cnri.reston.va.us>
References: <19991111080506.A1007@better.net> <14378.54900.714487.334690@weyr.cnri.reston.va.us>
Message-ID: <19991111120757.C1783@better.net>

On Thu, Nov 11, 1999 at 09:45:08AM -0500, Fred L. Drake, Jr. wrote:
> 
> William Park writes:
>  > 1.  Does anyone have experience with 'html2ps', 'html2tex', and
>  >     'html2latex' programs?  Or, is there any other program that will do
>  >     the conversion in Python or otherwise?  
> 
>   There's an html2ps.py in the Grail distribution; if your HTML
> doesn't involve tables that might just do it fo you.  You can even
> drive it from Python by using Grail's "printing" package.
>   Unfortunately, most HTML these days seems to involve at least tables 
> nested 12-deep.  ;-(

Found the solution by doing 'netscape -help'... duh!  See
    http://home.netscape.com/newsref/std/x-remote.html
Basically,
    $ netscape -remote "openURL(http://...)"
    $ netscape -remote "saveAs(test.ps,Postscript)"

William Park



From phd at phd.russ.ru  Mon Nov 15 11:24:26 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Mon, 15 Nov 1999 16:24:26 +0000 (GMT)
Subject: PyApache
In-Reply-To: <80p9uj$43f$1@oceanite.cybercable.fr>
Message-ID: <Pine.LNX.4.20.9911151623540.28300-100000@fep132.fep.ru>

Hi!

   There is PyApache mailing list, you should ask there.

On Mon, 15 Nov 1999, Florent Rami?re wrote:
>     I have the same problem !
>     Does anyone found a solution for this problem, or does pyApache
> absolutly do not handle such a functionnality (maybe) ?
> 
> Florent.
> 
> leon <leonar_n at epita.fr> wrote in message
> news:80p6p4$or5$1 at oceanite.cybercable.fr...
> > Hi,
> >
> > I'm using PyApache, and I would like to know if it's possible to make
> > 'persistent' modules. In fact I want to do only one "import file.py" like
> in
> > mod_perl, instead of loading this file each time the script is parsed.
> >
> > Thanks.

Oleg.
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From mcfletch at vrtelecom.com  Thu Nov 11 03:11:02 1999
From: mcfletch at vrtelecom.com (Mike Fletcher)
Date: Thu, 11 Nov 1999 03:11:02 -0500
Subject: Win32api.... can it create a Windows shortcut (.lnk) file?
In-Reply-To: <8KtW3.51128$Rx2.608380@news1.rdc1.bc.home.com>
Message-ID: <NDBBIAPHEKGOBIHCBLIICEOGCAAA.mcfletch@vrtelecom.com>

Here's a quicky demo of both creating and reading a lnk file on windows.
There's no real reason for the wrapper class other than that I'm a raving
lunatic and it's 3am :) ...

# link.py
# From a demo by Mark Hammond, corrupted by Mike Fletcher
from win32com.shell import shell
import pythoncom, os

class PyShortcut:
	def __init__( self ):
		self._base = pythoncom.CoCreateInstance(
			shell.CLSID_ShellLink, None,
			pythoncom.CLSCTX_INPROC_SERVER, shell.IID_IShellLink
		)
	def load( self, filename ):
		# Get an IPersist interface
		# which allows save/restore of object to/from files
		self._base.QueryInterface( pythoncom.IID_IPersistFile ).Load( filename )
	def save( self, filename ):
		self._base.QueryInterface( pythoncom.IID_IPersistFile ).Save( filename,
0 )
	def __getattr__( self, name ):
		if name != "_base":
			return getattr( self._base, name )

if __name__=='__main__':
	import sys
	file = sys.argv[1]
	shortcut = PyShortcut()
	if os.path.exists( file ):
		# load and dump info from file...
		shortcut.load( file )
		# now print data...
		print 'Shortcut in file %s to
file:\n\t%s\nArguments:\n\t%s\nDescription:\n\t%s\nWorking
Directory:\n\t%s'%(
			file,
			shortcut.GetPath(shell.SLGP_SHORTPATH)[0],
			shortcut.GetArguments(),
			shortcut.GetDescription(),
			shortcut.GetWorkingDirectory()
		)
	else:
		# create the shortcut using rest of args...
		data = map( None, sys.argv[2:], ("SetPath", "SetArguments",
"SetDescription", "SetWorkingDirectory") )
		for value, function in data:
			if value and function:
				# call function on each non-null value
				getattr( shortcut, function)( value )
		shortcut.save( file )

-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of me
Sent: November 11, 1999 1:54 AM
To: python-list at python.org
Subject: Win32api.... can it create a Windows shortcut (.lnk) file?


I'm just trying to find an easy to create shortcuts on Windows platforms
through Python... it doesn't seem to be a simple matter though.  I've gone
through all the Win32 docs, and found nothing, and have started writing
(with limited success) a C function to do it.

Is there an easy way?  q:]

Thanks,
Kevin Cazabon
kcazabon at home.com



--
http://www.python.org/mailman/listinfo/python-list




From aa8vb at yahoo.com  Mon Nov 29 08:20:45 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Mon, 29 Nov 1999 08:20:45 -0500
Subject: Newbie: raw_input and other input methods
In-Reply-To: <19991125.1481400@adam-aurag.penguinpowered.com>
References: <19991125.1481400@adam-aurag.penguinpowered.com>
Message-ID: <19991129082045.A1839643@vislab.epa.gov>

Hassan Aurag:
 |However when I wanted to have a nifty prompt like the following one:
 |
 |raw_input("\033[0;34mInput: \033[0m"), I was surprised it got all 
 |screwed up.
 |
 |This is true if you run this anywhere (gnome-terminal, vt without 
 |X...). It won't know the real size of terminal, because it thinks 
 |len("\033[0;34mInput: \033[0m") = 24 instead of what is really there 
 |and that is 7.
 |
 |This means that line wrapping won't work on your first line!
 |
 |My question is:
 |
 |Is there another input method a la raw_input (not like input that 
 |evals also) that can be mixed with readline and still sees that string 
 |as it is.

Well, an easy fix you probably thought of is to end your prompt string with
'\n'.  ;-)  That is, do input on the line following the prompt.

 |Or, is there any workaround?

Don't know.  A curses module may have what you need.  Too bad raw_input
doesn't support a line-length parameter, or you could just override the
default.

-- 
Randall Hopper
aa8vb at yahoo.com



From zakons at eccelerate.com  Wed Nov 10 09:22:52 1999
From: zakons at eccelerate.com (zakons at eccelerate.com)
Date: Wed, 10 Nov 1999 14:22:52 GMT
Subject: what is a function object?
References: <809p0a$8nf$1@nnrp1.deja.com> <m31z9znxy4.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <80bv3n$s9b$1@nnrp1.deja.com>

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

> it doesn't know a priori that the name "foo" is bound to a
> function. So it looks the name "foo" up, and assuming it finds
> something, attempts to call it with argument x. "Attempting to call
> it" goes something like:
>
> if type(foo) is types.MethodType:
>      apply(foo.im_func,(foo.im_self,x))
> elif type(foo) is types.FunctionType:
>      apply(foo,(x,))
> elif hasattr(foo,'__call__'):
>      apply(foo.__call__,(x,))

Thanks Michael. Your comment on the __call__ builtin method keyed me
into Mark Lutz's index to Programming in Python. Buried in the way
back on page 747, Closures in Python, he describes a class called
'counter' which implements __call__. What this does is make 'counter'
both an object and a method! I believe this is what it means when folks
say "functions are first class objects in python".

The real question is still "OK, so why would I want to make a method
(function) a full blown class (object)?" It seems like this makes sense
when there is some desired side effect every time the method is called.
(Is there a recognized pattern for this type of behavior?)

Stuart


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



From da at ski.org  Wed Nov  3 23:43:58 1999
From: da at ski.org (David Ascher)
Date: Wed, 3 Nov 1999 20:43:58 -0800 (Pacific Standard Time)
Subject: What are Aspects? (Re: How can I make my assertions smarter?)
In-Reply-To: <382150E0.F8938A7A@compaq.com>
Message-ID: <Pine.WNT.4.05.9911032042440.109-100000@david.ski.org>

On Thu, 4 Nov 1999, Greg Ewing wrote:

> William Tanksley wrote:
> > 
> > Aspects are really cool -- like a semantic preprocessor on steroids.
> 
> Where can we find out more about these Aspect thingies?

www.aspectj.org

--david





From ionel at psy.uva.nl  Mon Nov 29 06:46:36 1999
From: ionel at psy.uva.nl (Ionel Simionescu)
Date: Mon, 29 Nov 1999 12:46:36 +0100
Subject: Numeric write-in place facility
Message-ID: <81tovm$oe0@mail.psy.uva.nl>

Hi,

In a post on the wxPython list, Markus Gritsch suggested that array ops such
as
a = a+b
can be done faster by using write-in place:
Numric.add(a,b,a)

Playing around I see now that other ufuncs allow a write-in target
specification as their last argument. For example,

Numeric.sqrt(a,b)

will put the result in b (if its dimension and type are Ok).

The docs do not tell much about this.
Can anyone confirm/clarify/comment ?

Thanks,
ionel





From grant.griffin at iowegian.com  Mon Nov  8 17:40:15 1999
From: grant.griffin at iowegian.com (Grant Griffin)
Date: Mon, 08 Nov 1999 22:40:15 +0000
Subject: python: something completely different
References: <3820C0B8.62B2@iowegian.com> <38249ED2.BB7EA1DE@znet.com> <8028jv$7cf$1@hub.org> <38259F18.437@iowegian.com> <38278B5F.1AFF953C@python.net>
Message-ID: <3827514F.3A8F@iowegian.com>

Thomas A. Bryan wrote:
> 
> Grant Griffin wrote:
> 
> > I don't know if a Matlab-to-Python translator is currently available,
> > but if not, I don't think it would be all that hard to write one (in
> > Python, of course).
> 
> I'm sure how useful such a thing would be.  Most people I know who
> seriously use Matlab use at least one of its toolboxes. 

It would be a shame if somebody were to translate a whole toolbox. 
(Don't try this at home, kids.... ;-)

> They also
> tend to output plots/graphs when they're done with their work, and
> Matlab actually makes some very nice plots.  I haven't looked at the
> state of Python plotting/graphing in about a year, but when I last
> looked, I judged that the effort to produce nice plots with Python
> would be considerable.

One could take a cue from Octave here.  Octave provides Matlab-style
plot commands which are simply wrappers to Gnuplot.  (However, Octave's
reliance on Gnuplot creates a significant installation barrier in the
Windows world.)

Speaking as someone who is very new to Python, I am impressed by the
variety of plotting schemes available for it, as well as somewhat
bewildered by them.  From a new user's point of view, one would like to
be handed "standards" (while still having choices).  Python's use of Tk
is a good example of this balance relative to GUI.  I believe that the
inclusion of some sort of high-quality, portable, easy-to-install,
standardized plotting system would go a long way to helping Python
invade Matlab's turf.  Perhaps one of the plotting systems currently
available meet all these needs, if only it were given the blessing by
being distributed with NumPy.

> 
> That said, from what I remember of Matlab syntax, it probably would
> be a fairly straight-forward project to implement a translator for
> the basic vector/array/structure/scripting syntax (assuming that
> we're translating to a Python with Numeric Python compiled).  It
> might be a fun project, but it probably wouldn't move too many
> people away from Matlab until they could get colored, labeled,
> scaled plots from Python.  Most people would also have to be willing
> to code the few crucial routines that they've been using from a
> Matlab toolbox.  Of course, if the translator existed, maybe
> more people would try to move to Python.  That could spur a
> lot of growth in toolboxes for Numeric Python! :)

I guess the distinction is that Matlab really is a "system" (consisting
of a language, toolboxes, and plotting), whereas Python is primarily a
"language", which obviously is extensible _into_ a system (many systems,
actually).  Therefore, Matlab users aren't going to switch in large
numbers until a Python alternative becomes viable as system.  OTOH,
since Python would give them many advantages outside of the Matlab
paradigm (as well as being free/open), that's an inducement.

Python's extensibility already allows it to easily "co-opt" C code to
form new libraries, but the existance of a high-quality Matlab-to-Python
translator would largely solve the library problem, since existing
free/open libraries from Octave, Scilab, et.al. could be automatically
translated.  The most obvious way to write a translator like this would
be in Python itself, but perhaps the most expedient way to write a truly
robust translator would be to borrow the parser/lexer from Octave; given
a tree of Matlab syntax generated by the guts of Octave, one could then
"pretty-print" the tree into Python.  (Sounds like a fun project for you
language hobbyists out there...)

Last, since several plotting systems are already available for Python,
one could be chosen as a standard, then adapted as necessary to provide
functionality similar to Matlab's plots.

=g2
-- 
_____________________________________________________________________

Grant R. Griffin                                       g2 at dspguru.com
Publisher of dspGuru                           http://www.dspguru.com
Iowegian International Corporation	      http://www.iowegian.com



From gerrit.holl at pobox.com  Tue Nov 30 12:00:38 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Tue, 30 Nov 1999 18:00:38 +0100
Subject: Htmlizing text
In-Reply-To: <005b01bf3b3b$1c14d5e0$f29b12c2@secret.pythonware.com>; from fredrik@pythonware.com on Tue, Nov 30, 1999 at 02:59:23PM +0100
References: <943909426snz@vision25.demon.co.uk> <19991129170705.A314@better.net> <19991130141224.A2813@stopcontact.palga.uucp> <002301bf3b39$5b5bc9e0$f29b12c2@secret.pythonware.com> <005b01bf3b3b$1c14d5e0$f29b12c2@secret.pythonware.com>
Message-ID: <19991130180038.A2594@stopcontact.palga.uucp>

Fredrik Lundh wrote:
> I wrote:
> 
> > Gerrit Holl <gerrit.holl at pobox.com> wrote:
> > > By the way, why doesn't the cgi module does this?
> > 
> > if you use ISO Latin 1 for your pages (which happens to
> > be default in HTTP/1.1, HTML 3.0, and real life), you
> > don't really have to quote all possible entities just be-
> > cause you can...
> 
> and yes, your code assumes that the original string
> always uses ISO Latin 1.  the cgi version works fine
> with plain ASCII, and *any* 8-bit ISO encoding.

My code does, but one could also write code that doesn't. Code that
converts platform-depentent HTML code to platform-indepentent HTML
code.

regards,
Gerrit.

-- 
"Calling EMACS an editor is like calling the Earth a hunk of dirt."

  -- Chris DiBona on Dirt (Open Sources, 1999 O'Reilly and Associates)
  5:59pm  up  2:15, 10 users,  load average: 1.32, 1.35, 1.18



From skip at mojam.com  Mon Nov 22 10:27:09 1999
From: skip at mojam.com (Skip Montanaro)
Date: Mon, 22 Nov 1999 09:27:09 -0600 (CST)
Subject: Order of keyword arguments
In-Reply-To: <14392.20852.400630.148814@dolphin.mojam.com>
References: <38379233.ADE65CDB@corrada.com>
	<19991121034245.A21255@quark.emich.edu>
	<38380F92.D9817451@corrada.com>
	<14392.20852.400630.148814@dolphin.mojam.com>
Message-ID: <14393.24781.600654.739477@dolphin.mojam.com>

Reading my post from the weekend, it seems to me I wasn't very clear about
things.  I wrote:

    Skip> I solved this in one instance by giving all the input tag names
    Skip> names like XXXtagname, where XXX is a zero-padded number.  The
    Skip> order of display is determined by the form's author, so you could
    Skip> have

    Skip>     001name
    Skip>     002address
    Skip>     003address
    Skip>     004city
    Skip>     005state

    ...

What I meant was that the HTML form's author could determine the sort order
of the input arguments so that further display of them would be correct,
regardless of the order in which they were passed to the cgi script.

Skip Montanaro | http://www.mojam.com/
skip at mojam.com | http://www.musi-cal.com/
847-971-7098   | Python: Programming the way Guido indented...



From aahz at netcom.com  Wed Nov 10 12:05:44 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 10 Nov 1999 17:05:44 GMT
Subject: [Python] header (fwd)
References: <Pine.HPX.4.20.9911081711570.27637-100000@ztt.rsn.hp.com> <807lbm$ail$1@nntp5.atl.mindspring.net>
Message-ID: <80c8l8$r0$1@nntp9.atl.mindspring.net>

In article <807lbm$ail$1 at nntp5.atl.mindspring.net>,
Aahz Maruch <aahz at netcom.com> wrote:
>In article <Pine.HPX.4.20.9911081711570.27637-100000 at ztt.rsn.hp.com>,
>Benoit Ciceron  <bciceron at rsn.hp.com> wrote:
>>
>>can we get an automatic subject header : "[Python]" to ALL messages
>>submitted to the list , including replies and newsgroup threads ?
>
>Why?

Following up on my own post, I got a response from Benoit in e-mail that
appeared to have been cc'd to the Python mailing list -- but it hasn't
shown up on the newsgroup.  That illustrates my point precisely: the
Python mailing list is gated to a Usenet newsgroup and therefore should
have as little mucking as possible.
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From josh at linc.cis.upenn.edu  Thu Nov  4 15:50:17 1999
From: josh at linc.cis.upenn.edu (Joshua Marcus)
Date: 4 Nov 1999 20:50:17 GMT
Subject: file i/o !!! help
References: <7vsolo$aer$1@usenet01.srv.cis.pitt.edu>
Message-ID: <7vsri9$fi9$1@netnews.upenn.edu>

Here's one way to do it:

import string

DELIMETER = " "

fp = open("foo.txt", "r") # open file for reading
floating_point_list = []  # initialize list
while 1: # loop til break
   first_line = fp.readline() # put the next line
                              # in first_line
   if first_line == "":       # if the file is done,
        break                 # break
   # the following breaks down the line "1 2 3 4" to
   # the list ["1", "2", "3", "4"]
   string_list = string.split(first_line, DELIMETER)
   for number_string in string_list:
       # string.atof converts the string to a float
       # floating_point_list.append(x) appends x to the
         list floating_point_list
       floating_point_list.append(string.atof(number_string))

print "%s" % floating_point_list

fp.close()

--j
tdoe (tcdoe at pitt.edu) wrote:
: Hi,
: im very new to using python, and im having a TERRIBLE time trying to find
: some example scripts for FILE I/O !!!!
: all i want to do is read a "comma" or "space" delimited list of numbers from
: a file such as:

: 1 2 3 4
: 5 6 7 8
: 9 10 11 12

: and input them into a floating point list variable.

: well, this is probably a "cinch", but
: for me this has been tough.

: any help very appriciated.
: thanks,
: todd
: tcdoe at pitt.edu






From arnold at dstc.edu.au  Sun Nov 28 17:41:38 1999
From: arnold at dstc.edu.au (David Arnold)
Date: Mon, 29 Nov 1999 08:41:38 +1000
Subject: status of remote debugging? 
In-Reply-To: Your message of "Fri, 26 Nov 1999 16:56:13 GMT."
             <81me3c$j51$1@nnrp1.deja.com> 
Message-ID: <199911282241.IAA30140@piglet.dstc.edu.au>

-->"Preston" == Preston Landers <prestonlanders at my-deja.com> writes:

  Preston> Hello all, Just wondering if there is already an
  Preston> implementation of remote debugging for Python.  If not, I'm
  Preston> thinking it would be a neat trick.

several years ago, Martin Chilvers (of Fnorb fame) hacked up a Tk
interface (i think) which used a locally-developed notification
service to communicate with the target program.

it was very nice -- full control of the remote program, but it did
require threads in the target.  martin?  do you still have it?


more recently, i was talking to Mark Hammond about the debugging stuff
Microsoft has done for its scripting languages: they have decreed a
set of interfaces to be provided by the interpreter, making the
internal state visible to a debugger process.  this has the added
benefit of support inter-language calls.

i've been toying with the idea of doing something similar, with bits
on the side of pdb and maybe even gdb(?!) to expose the runtime state
in some uniform way.  but i haven't actually done anything yet ...




d



From mal at lemburg.com  Sat Nov  6 05:59:19 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Sat, 06 Nov 1999 11:59:19 +0100
Subject: pybench...where ?
References: <19991106113537.A17850@sz-sb.de>
Message-ID: <38240A07.9086F5E5@lemburg.com>

Andreas Jung wrote:
> 
> I remember that there has been a PyBench suite available
> some time ago. However I can not find the current version.
> 
> Any hint ?

	http://starship.skyport.net/~lemburg/pybench-0.6.zip

-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                    55 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/





From sp00fD at yahoo.com  Wed Nov 24 15:26:52 1999
From: sp00fD at yahoo.com (sp00fD)
Date: Wed, 24 Nov 1999 20:26:52 GMT
Subject: mysql and DATE format
Message-ID: <81hhm6$ba2$1@nnrp1.deja.com>

I've currently got some perl code that executes a sql statment which
looks like:

SELECT id, subject, author, text, DATE_FORMAT(date, \"%W, %d %b %Y\"),
DATE_FORMAT(date, \"%h:%i %p\") FROM Table WHERE thread=18 ORDER by
date desc

When trying to use that in python like this:
"""SELECT .. DATE... FROM %s WHERE thread=%d""" % (table, id)

it tries to expand the DATE %W.. variables.
I've also tried it as

sql = 'SELECT id, subject, author, text, '
sql = sql + 'DATE_FORMAT(date, \"%W, %d %b %Y\"), '
sql = sql + 'DATE_FORMAT(date, \"%h:%i %p\") FROM' + table
sql = sql + 'WHERE thread=' + id + 'ORD...'

which gives me a TypeError: illegal argument type for built-in
operation.  Why?  How can I do this properly?


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



From smst at quantisci.co.uk  Tue Nov 16 04:54:57 1999
From: smst at quantisci.co.uk (Steve M S Tregidgo)
Date: Tue, 16 Nov 1999 09:54:57 GMT
Subject: Memory not being released?
Message-ID: <199911160954.JAA12700@cobweb.quantisci.co.uk>

Thanks for your response.  Some more of my own comments in reply:

gmcm at hypernet.com wrote:
> Steve Tregidgo writes:
> 
> > The actual data that results from the process is relatively small
> > (the useful stuff is less than 1MB, and other data -- such as
> > objects that have been dealt with, to prevent following circular
> > references -- is no more than a few MBs), and certainly well
> > below the size of the actual footprint.
> >
> > On the other hand, commenting out some other expressions (such as
> > those that remember data -- in other words, the things that I
> > would expect to take up memory and not release it again) doesn't
> > make a blind bit of difference.
> 
> That's a pretty good sign that you have circular references that 
> are keeping objects alive. Check out Cyclops from 
> ftp://ftp.python.org/pub/python/contrib/System/.
> 

Thanks, I'm looking at those file right now.

> Also, be aware that even if your python objects are getting 
> collected, that doesn't mean your c runtime is giving the 
> memory back to the OS. Although 250M would indicate a 
> pretty horrid c runtime.
>  
> > So what's going on?  My guesses so far have been along the lines
> > of refcounts not going down -- how else do I explain why memory
> > that was allocated was seemingly not deallocated?  I've tried
> > del'ing everything in sight, storing ids instead of objects,
> > inserting the odd sys.exc_traceback=None to clear those pesky
> > traceback reference holders ... none of it has helped so far.
> > 
> > Are there other hidden things that might hold references to my
> > objects?  And if so, what can I do about them?  After using an
> > object once, I don't want (don't need) to keep it in memory any
> > more, so for those objects I'll be revisiting at a later date, I
> > just remember their id and delete them.
> 
> In your code, dict1 is holding a reference to every object 
> you've visited, and thus keeping them all alive.

I'm storing obj.__id__ in each case, which is an integer -- I
certainly hope that it doesn't count as a reference to the object. 
Does it?  In mock-ups of the routine, storing a big list or
dictionary or numbers, the few MBs taken up by those structures were
not on a par with the size of the process in question.  Finally,
removing the lines (in the actual code, not the mock-up) which store
the dictionary didn't lower the footprint.  Frustrating, to say the
least!

> OTOH, judging 
> by "for item in obj.subs:", it looks like the whole graph is alive 
> all the time anyway. Essentially, you've got a huge pointer-
> based data structure. You need to hoist most of the smarts up 
> to a Graph object, and use some kind of persistent key to 
> identify objects (which are really on disk, or recreated on the 
> fly...). So that snippet might be:
>   for itemname in obj.subs:
>     obj = graph.getobject(itemname)
>  

This is something I considered -- at one stage in my attempts at
fixing the problem, I created a list of id's (so item.__id__ for each
item in obj.subs), deleted the copy of obj.subs (I was actually using
a list created by a method of obj, not an actual attribute), and then
retrieved a persistent item at the start of the loop, deleting it at
the end.  We do have persistent keys for objects, so I'm going to
take a look at this possibility again...

> > If it's any help, the process does something like this:
> > 
> > def recurse(obj, list, dict1, dict2):
> >   # Prevent circular recursion...
> >   if dict1.has_key(obj.__id__):
> >     return
> >   else:
> >     dict1[obj.__id__] = None
> > 
> >   a = do_thing(obj, dict2)
> > 
> >   # Recurse into obj's children, or do something else
> >   for item in obj.subs:
> >     if item.spam:
> >       recurse(item, list, dict1, dict2)
> >     else:
> >       b = do_thing(item, dict2)
> > 
> >   c = do_thing(obj, dict2)
> > 
> >   # Remember some things -- commenting
> >   # out produces no memory saving.
> >   list.append(a,b,c)
> > 
> > The function do_thing implements a loop; in total (over the whole
> > process) the loop's body is executed tens of thousands of times
> > -- it causes the biggest memory saving when commented out (and of
> > course it's the most important bit), but does no obvious
> > "remembering" of things.
> > 
> > 
> > Cheers,
> > Steve
> > 
> > --
> >        -- Steve Tregidgo --
> > Developer for Business Collaborator
> >         www.enviros.com/bc
> > 
> > 
> > Sent via Deja.com http://www.deja.com/
> > Before you buy.
> > 
> > -- 
> > http://www.python.org/mailman/listinfo/python-list
> 
> 
> 
> - Gordon



From a.eyre at optichrome.com  Tue Nov 30 06:02:42 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Tue, 30 Nov 1999 11:02:42 -0000
Subject: difference between invocation of python script on unix and microsoft?
In-Reply-To: <81vkig$228@news.service.uci.edu>
Message-ID: <001501bf3b22$6c3f0720$3acbd9c2@peridot.optichrome.com>

> Can anyone tell me in what ways using python under microsoft is
> different from python under unix?

If you have associated .py files with python.exe, then you can do:

myfile.py < myinput.txt

Otherwise it's:

python myfile.py < myinput.txt

Although I did find with MKS Toolkit installed, if the first line is:

#!/usr/bin/env python

Then you can name the script 'myfile' and do:

. myfile < myinput

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From NOSPAMbruce_wheeler at hp.com.NOSPAM  Wed Nov 10 16:18:53 1999
From: NOSPAMbruce_wheeler at hp.com.NOSPAM (Bruce Wheeler)
Date: Wed, 10 Nov 1999 13:18:53 -0800
Subject: .dll and .exe file version informaton
References: <000101bf2b3c$a7d43e40$d32d153f@tim> <0E16861EE7BCD111BE9400805FE6841F0C7A489C@c1s5x001.cor.srvfarm.origin-it.com>
Message-ID: <3829E13D.23E99B91@hp.com.NOSPAM>

Robin Boerdijk wrote:
> 
> Tim Peters <tim_one at email.msn.com> wrote in message
> news:000101bf2b3c$a7d43e40$d32d153f at tim...
> > [emuller at painewebber.com]
> > > Can anyone give me an idea how to quikly get the version info on .dll
> > > and .exe files under windows?
> >
> > In Explorer, right-click on the file, select Properties from the context
> > menu, and click the Version tab.
> 
> Strange... My Explorer only gives me a General tab. Quickview in the context
> menu, however, shows a lot of technical info about the selected .dll or
> .exe.
> 
> Robin

Not all .dll or .exe contain version info. Perl.exe does, python.exe
doesn't although it is in python15.dll. You need to include the info in
a resource file. If you want to get at it from a program, use the win32
api - GetFileVersionInfo. Possibly this is available for Python in Mark
Harmmond's win32all package.

regards,
Bruce_Wheeler at hp.com



From Vladimir.Marangozov at inrialpes.fr  Mon Nov 29 01:20:50 1999
From: Vladimir.Marangozov at inrialpes.fr (Vladimir Marangozov)
Date: Mon, 29 Nov 1999 07:20:50 +0100
Subject: creating an AIX shared object for Python
References: <81haen$5qg$1@nnrp1.deja.com>
Message-ID: <38421B42.E847BF8F@inrialpes.fr>

drjatalb at my-deja.com wrote:
> 
> [snip]
> 
> cc -c -qfuncsect -I/users/toollib/include/python1.5 -
> I/net/rush/disk1/pdi/toolkits/crusher/crusher312/include crushermodule.c
> 
> ld -G crushermodule.o -o crushermodule.so -
> L/net/rush/disk1/pdi/toolkits/crusher/crusher312/lib/AIX_4.2.1 -lcrusher
> ld: 0711-327 WARNING: Entry point not found: __start
> ld: 0711-244 ERROR: No csects or exported symbols have been saved.
> 
> Any suggestions would be appreciated.
> 
> Dr. Jeff Talbert
> jat at iti-oh.com

Instead of ld -G, use the ld wrapper ld_so_aix. See Misc/AIX-NOTES in
the distribution for more details.

-- 
       Vladimir MARANGOZOV          | Vladimir.Marangozov at inrialpes.fr
http://sirac.inrialpes.fr/~marangoz | tel:(+33-4)76615277 fax:76615252



From bwarsaw at cnri.reston.va.us  Tue Nov  2 17:16:03 1999
From: bwarsaw at cnri.reston.va.us (Barry A. Warsaw)
Date: Tue, 2 Nov 1999 17:16:03 -0500 (EST)
Subject: pygettext
References: <7vnj5l$ha$1@essle.valdyas.org>
Message-ID: <14367.25251.866834.970374@anthem.cnri.reston.va.us>

>>>>> "BR" == Boudewijn Rempt <boud at rempt.xs4all.nl> writes:

    BR> I'm trying to write a resources page to go with my Python/KDE
    BR> tutorial, and I want to point people to pygettext.py, which I
    BR> found somewhere on my harddisk. However, there isn't a
    BR> copyright notice in the file, and I'd like to mention the
    BR> author, too. Could he please contact me?  I've had to adapt
    BR> the program slightly to work with the kde i18n function, too.

I wrote the original version, and while a few folks have sent me
patches, haven't had much time to develop it further.  You can get to
it from

    http://www.python.org/~bwarsaw/software/pyware.html

And please do feel free to send me changes.  Eventually I'll make a
new release.

-Barry



From gvwilson at nevex.com  Mon Nov 29 09:02:13 1999
From: gvwilson at nevex.com (Greg Wilson)
Date: Mon, 29 Nov 1999 09:02:13 -0500
Subject: XML tool status?
Message-ID: <DIv04.8846$cz3.320849@nnrp1.uunet.ca>

Hi.  I've gone through the Python/XML HOWTO, and would like to start playing
with the DOM module it describes, but:

1) I can't find a Windows version of the software; and

2) the Python XML SIG status page has a "most recent update" date from
December 1998 --- is that fossilware, or has there really not been anything
new posted since then?

Thanks,
Greg






From thomas at xs4all.nl  Tue Nov  2 09:16:57 1999
From: thomas at xs4all.nl (Thomas Wouters)
Date: Tue, 2 Nov 1999 15:16:57 +0100
Subject: General q. `bout subclassing
In-Reply-To: <381EE780.EBF18F59@bibsyst.no>; from thomas@bibsyst.no on Tue, Nov 02, 1999 at 02:30:41PM +0100
References: <381EE780.EBF18F59@bibsyst.no>
Message-ID: <19991102151657.G29354@xs4all.nl>

On Tue, Nov 02, 1999 at 02:30:41PM +0100, Thomas Weholt wrote:

> Simple one :

> How do I subclass fundamental stuff like lists etc,

You dont.

> tried something like :
> 
> > >>> class my_list([]):
> > ...     a = ''
> > ...
> > Traceback (innermost last):
> >   File "", line 1, in ?
> > TypeError: base is not a class object
> > >>>
> >

> Huh ??

built-in types, and types provided by C extention modules (as opposed to
Python modules) are not 'full' objects in the sense that Python objects are
-- they do not have a __dict__ attribute listening all attributes, and they
do not have a __class__ attribute pointing to the 'parent' class. You also
can't subclass them. Instead, you have to wrap them in a base class, and
then extend them:


Class MyList:
	def __init__(self, orig=None):
		self.reallist = (orig or [])[:]
	def __getitem__(self, index):
		return self.reallist[index]
	def __setitem__(self, index, value):
		self.reallist[index] = value
<etc>

Class MySubclassedList(MyList):
	<etc>

The standard Python library provides such wrappers, and do their best to
simulate built-in types. They do a pretty good job, too ;-) Check out
the 'UserList' and 'UserDict' modules -- they are, by necessity, written in
python. Very readable :-)

I believe the classless-builtin-types problem is supposed to be fixed by
Python 2.0. But I've seen that in response to so many ideas... Either
python2 is going to be as bloated as C++, or it is a generic 'when hell
freezes over on a midsummer christmas easter-wednesday'-type answer. :)

python-forever'ly y'rs,
-- 
Thomas Wouters <thomas at xs4all.net>

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



From hamish_lawson at my-deja.com  Mon Nov 15 06:15:50 1999
From: hamish_lawson at my-deja.com (hamish_lawson at my-deja.com)
Date: Mon, 15 Nov 1999 11:15:50 GMT
Subject: Python module to format text into columns?
Message-ID: <80oq13$ql3$1@nnrp1.deja.com>

Is there a Python module that will take a list of
strings and wordwrap them columnwise, similar to
Perl's format mechanism? For example, if supplied
with the following strings

   "14/11/1999"
   "Here is a a paragraph of text."
   "Here is another paragraph of text."

and with appropriate specification of column
widths and alignments, it could produce something
like:

   14/11/1999   Here is a   Here is another
                paragraph   paragraph of text.
                of text.


Hamish Lawson


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



From mwh21 at cam.ac.uk  Sun Nov 14 05:16:05 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 14 Nov 1999 10:16:05 +0000
Subject: GNU readline module
References: <80lcsk$99g$1@watserv3.uwaterloo.ca>
Message-ID: <m3puxdpee2.fsf@atrus.jesus.cam.ac.uk>

kvijayan at calum.csclub.uwaterloo.ca (Kannan Vijayan) writes:

> hi
> 
> I'm trying to get the readline module
> to work in python.
> I have an app to which I want to add
> tab-completion of commands and other things.
> 
> The docs for the GNU readline module are
> extremely sparse.  

Yes.

> Most of what I know about the module are derived from looking at
> it's source, looking at how "rlcompleter" uses it, and my own sparse
> knowledge of the C readline implementation itself.

I've generally been able to make it do what I want. Not always easily
(exceptions getting binned can be really confusing).

> As far as I can see, there is no hook to the
> "readline(char *PROMPT)" function in the
> readline module.  Am I missing something?

That's raw_input, more or less.
 
> I have, however, gotten my completer function
> to work well from within the interactive
> interpreter environment.  What I'm having
> trouble with is getting it to work from a
> non-interactively interpreted application in
> it's own happy .py file.  I assume there would
> be some function which I could call to read a
> line, which in turn would call my completer
> function to do the job when a tab is pressed.

Are you aware that to activate readline away from the interactive
interpreter you have to import readline explicitly? Chuck an 

import readline

at the top of your main script and give it another go.

> I have a nagging notion that maybe the readline
> module is only for use from withing the
> interactive interpreter.  IS this true? if so..
> that's extremely dissapointing.

I've never seen it used for anything else. There's no particular
reason for this. I don't know whether it's inclusion was motivated by
enhancing the interactive environment or allowing it's use it other
applications. Any PyHistorians care to comment?

Regards,
Michael



From fredrik at pythonware.com  Mon Nov  1 11:46:52 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 1 Nov 1999 17:46:52 +0100
Subject: Listbox.. ACTIVE-item is the one selected previously..
References: <ksemebjifu.fsf@telelogic.se>
Message-ID: <009301bf2488$b3c50910$f29b12c2@secret.pythonware.com>

Andrew Markebo <andrew.markebo at telelogic.se> (or is it Eskilsson?) wrote:
> I want to have a listbox, and when the user clicks on one item in it,
> he gets the value selected 'published' somewhere.. with the small
> program included below.. If it is launched, and the user selects the
> letters in order I would assume the result to be "abcde", but what I get is:
> "aabcd", after the first selection, ACTIVE is lagged.. 
> 
> What have I missed??

by default, instance bindings are called *before* class
bindings -- and it's the class bindings that implement
standard behaviour.

easy solution: use double click instead.

slightly harder solution: use bindtags to swap
instance and class bindings.

best solution: instead of using instance bindings,
create your own binding class, and place it be-
fore the standard class bindings.

(sorry, no time for code samples today).

</F>




From fredrik at pythonware.com  Fri Nov 19 07:54:32 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 19 Nov 1999 13:54:32 +0100
Subject: The First Annual Obfuscated Python Contest
References: <mfbZ3.71$AZ6.1807@dfiatx1-snr1.gtei.net>
Message-ID: <001d01bf328d$3ac77170$f29b12c2@secret.pythonware.com>

Jules Allen wrote:
> I'm kidding, of course.

yeah, calling it "first" would be a great lie ;-)

for more masterpieces, see:

    http://starship.python.net/crew/amk/writing/crypto-curiosa.html
    (from where the slashdot item was stolen)

    http://www.python.org/doc/FAQ.html#4.15

and greg's all-time masterpiece:

http://www.deja.com/getdoc.xp?AN=398922825&fmt=text
(make sure to place it all on one line)

</F>




































here's the solution to greg's code; don't look at it
unless you've spent at least 15 minutes on it your-
self:

http://www.deja.com/getdoc.xp?AN=398962967&fmt=text





From cgw at fnal.gov  Fri Nov  5 09:24:44 1999
From: cgw at fnal.gov (Charles G Waldman)
Date: Fri, 5 Nov 1999 15:24:44 +0100 (MET)
Subject: Drawing (directed) graphs
In-Reply-To: <y0jeme5owc9.fsf@vier.idi.ntnu.no>
References: <000a01bf2792$b83ff150$3acbd9c2@peridot.optichrome.com>
	<y0jeme5owc9.fsf@vier.idi.ntnu.no>
Message-ID: <14370.59564.979979.285996@gargle.gargle.HOWL>

Magnus L. Hetland writes:
 > "Adrian Eyre" <a.eyre at optichrome.com> writes:
 > 
 > > > Good thing there is a platform independent Python graphics library!
 > > 
 > > How can there be. Some OSs (Minix) don't support graphics at all...
 > 
 > Maybe you should check out the URL given before whining <wink>
 > 
 > Even on Minix you could easily generate jpegs, or postscript files or
 > whatever with Piddle. If you can't run an X server or something, that
 > only means that you can't see it... Or maybe you have a printer?

Or for maximum portatbility, you could use the "aa" ascii-art library
to display the images on an ASCII terminal.... it's a C library, but
it would be trivial to SWIG up a set of Python bindings, if this
hasn't been done yet.




From JamesL at Lugoj.Com  Thu Nov 25 19:56:12 1999
From: JamesL at Lugoj.Com (James Logajan)
Date: Thu, 25 Nov 1999 16:56:12 -0800
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk>
Message-ID: <383DDAAC.DB9F6CE2@Lugoj.Com>

Mark Carroll wrote:
> Basically, I'd be really interested to learn: what _don't_ people like
> about Python?

Well, I guess I don't like all the complaints about Python, not that there
all that many. And I don't like the fact that Guido spelled "define" wrong.
He incorrectly spelled it "def". This has annoyed me for years and I'm not
sure I'll continue to use Python until this gets corrected. I've also
considered no longer using C and Unix for years because K&R incorrectly
spelled "create" as "creat".

"F u cn rd ths, thn u cn bcm a prgrmmr!"

(Too cheap to buy a vowel and too lazy to be taken seriously.)



From robin at jessikat.demon.co.uk  Tue Nov  2 08:31:33 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Tue, 2 Nov 1999 13:31:33 +0000
Subject: embedding stubbed tcl
Message-ID: <bLd7vSA1euH4EwDS@jessikat.demon.co.uk>

Python uses tkinter which embeds tcl and tk. At present python 1.5.2
comes with a dll that explicitly links to tcl80.dll. having just moved
to the latest tcl8.2.1 I have to recompile this dll and with a very few
minor changes all works ok. What happens when 8.3 appears? I guess I
have to do this again :(

What is the simplest way to avoid this pain using stubs? I mean what are
the guaranteed entry points in the dlls. I need the interp pointer
before I can start off the stubs interface and it seems that the first
thing done by tkinter does is Tcl_CreateInterp followed by
TclpInitLibraryPath(baseName). The latter appears not to be in the stubs
library. Therefore it seems as though tkinter is not stubs friendly. I
guess that the embedding structure for tkinter is a bit out of sync with
the correct way of doing things.
-- 
Robin Becker



From paul at prescod.net  Wed Nov  3 11:46:42 1999
From: paul at prescod.net (Paul Prescod)
Date: Wed, 03 Nov 1999 10:46:42 -0600
Subject: Comparison between Python and "Ruby"
References: <1270663667-24391282@hypernet.com> <381ED7A6.D7D7FB13@compaq.com> <877lk1wiij.fsf@ev.netlab.co.jp> <19991102154543.A4469@optiplex.palga.uucp>  <381F9EB4.63FA033F@python.net> <slrn81ve6m.l1c.neelk@brick.cswv.com>
Message-ID: <382066F2.D12ECDA2@prescod.net>

Neel Krishnaswami wrote:
> 
> Python (and Ruby, for that matter) remind me more of Smalltalk than
> anything else -- ie, they have extensive libraries, a community with a
> love of elegance but not of purity[*], and a greater focus on practical
> applications than on academic research.

Isn't the Smalltalk community pretty enamored of purity? I mean "if" is
an object that takes messages, right?

 Paul Prescod



From gmcm at hypernet.com  Sun Nov  7 11:32:07 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sun, 7 Nov 1999 11:32:07 -0500
Subject: __str__ vs. __repr__
In-Reply-To: <u4glOIWXzGrqvJnI+IlPAQeurz1J@4ax.com>
Message-ID: <1270130172-9094859@hypernet.com>

Toby Dickenson writes:

> "Tim Peters" <tim_one at email.msn.com> wrote:
> 
> >Let me back off to what repr and str "should do":
> >
> >repr(obj) should return a string such that
> >    eval(repr(obj)) == obj
> 
> I don't understand the motivation for this requirement. Why would
> anyone want to pass such a string to eval? If you anticipate the
> need for reconstructing the object from a textual representation,
> then surely pickle is a better option?

It's not human editable. I guess you've never used 
 open('xxx.xxx','w').write(repr(dict))
and
 dict = eval(open('xxx.xxx','r'),read())

as a cheap config file.

> OK, so repr as it stands today is clearly not suitable for
> formatting for people, and is less than ideal for formatting for
> the machine. 

For strings, ints, lists, tuples and dicts it is dandy for both 
right now. For longs it's good for machines; for some others 
it's OK for some users.

It doesn't make a whole lot of sense for low-level objects 
involving open resources, but it does make a great deal of 
sense for user objects which would represent a URL as a 
string, or a file as a (filename, mode). You can then make 
repr(object) yield a valid constructor and include high level 
objects in your config file, or cut and paste from interactive to 
a script.

Whether you use them or not, these are very valuable features 
that are promised (but not entirely fulfilled). 

> However there is a third audience for who it is
> ideal, the programmer, who need a representation for use in
> development tools such as debuggers, or the interactive 
mode.
> 
There's been some debate about interactive mode. I would 
personally prefer that interactive mode err on the side of 
excess accuracy, (I find pprint to be a necessity for complex 
structures, and whether repr or str is used, that's not going to 
change). Debuggers should not be relying on either, but 
should pick apart the object through knowledge of Python 
internals.

> Tim's Rat class is a good example of how __repr__ should not be
> implemented, because it exposes the internal representation of
> the object, not the external value.

Tim's Rat class should provide a separate __str__ 
implementation.
 
> However, using a small number of digits for repr(float) does make
> sense because it still provides sufficient precision for most
> programmers.

For human consumption, I *always* use % formatting with 
floats - there's no way that float can know whether the human 
is a chemist or a bean-counter, (if the latter qualify at all).


- Gordon



From aahz at netcom.com  Mon Nov 29 18:37:09 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 29 Nov 1999 23:37:09 GMT
Subject: Python meta object question
References: <3842D025.3283C871@linux-france.org>
Message-ID: <81v2n5$kjf$1@nntp9.atl.mindspring.net>

In article <3842D025.3283C871 at linux-france.org>,
Mickael Remond  <mikl at linux-france.org> wrote:
>
>Is there a way to automatically update the class tree to show the change
>in inherited class attributes after a class redefinition ?

The problem is that the original class definition still has references
to it as long as there are instances of that class.  Your problem is
that you're trying to make inheritance do the work of object
composition; a much better way to handle this specific problem is to
create an instance of a company object, set self.companyName, then pass
that company object to each of the employees.  When you update that
central company instance, each employee instance will automatically
access the new information.
--
                      --- Aahz (@netcom.com)

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

Question authority: ask me anything!



From donn at u.washington.edu  Wed Nov 10 12:06:44 1999
From: donn at u.washington.edu (Donn Cave)
Date: 10 Nov 1999 17:06:44 GMT
Subject: Text Manipulation in Python
References: <VA.00000012.055a8702@ops.corpex.com>
Message-ID: <80c8n4$1656$1@nntp6.u.washington.edu>

Quoth Edward Hasted <edward at corpex.com>:
| We want to use it to alter specific lines in template files, typically 
| something like changing:-
|
| Variable = 1234
|
| to 
|
| Variable = Company Name
|
| The text manipulation strings within Python seem to work sequentially 
| rather than on a line basis.
|
| 1. Is this correct?

Hard to say, that analysis may be a little too terse.

| 2. What is the best way to do text searching and manipulation within 
| Python.

Depends!  I thought the system Thomas Fuchs proposed in his followup
was interesting, and if it works for you, perhaps that's close enough
to the best - but there isn't enough detail here to know whether it's
suited to your needs at all.

Another followup mentioned that perhaps the best way is to not do it
in Python, and there's something to that too.  Certainly on UNIX, or
platforms like BeOS that borrow from its repertoire, there are an
embarrassment of riches in text manipulation languages, and it's a
point of tradition that the language or tool to solve each little
problem be considered separately (I don't mean that to sound silly,
either, but will not expound further here.)

In a project of mine, on BeOS, I've been using "m4", a macro language,
to do what you seem to have in mind.  Most of the heavy lifting is done
in Python, but the result goes into a template processed by m4.  It's
kind of like the C preprocessor but much more capable, and despite its
relative obscurity, usually it's already there on a normal UNIX platform,
in my experience anyway.

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu



From damien_morton at my-deja.com  Mon Nov 15 01:12:55 1999
From: damien_morton at my-deja.com (damien_morton at my-deja.com)
Date: Mon, 15 Nov 1999 06:12:55 GMT
Subject: Production Web Deployment?
References: <199911080604.BAA22580@python.org> <NBBBKMMANMGEEBHPJJECGEGIFJAA.dirkl@home.com> <14376.24392.828292.926057@dolphin.mojam.com>
Message-ID: <80o895$ebn$1@nnrp1.deja.com>

Im working at a small publishing outfit, and we've been pumping out 80
million hits a month for www.maximmag.com, from a 266MHz IIS/ASP box.
Weve just moved onto a 3-way clustering server, and expect to be able
to take pretty much anything that can happen. From time to time, our
site get mentioned on TV or by Howard Stern and it goes berserk, but we
expect to be able to handle that too.

All of our software is written by people without much regard for
performance or optimisation (e.g. stupidly complex SQL queries feature
in almost every page).

Ive considered writing caching software within ASP, which would cache
the semi-dynamic parts of the website (almost all of it), but its looks
like as long as our populaity grows at something less than Moore's law,
we can pretty much solve any problems by throwing more or better
hardware at it.

Ive looked long and hard at Zope and pretty much ANY Python oriented
solution which might be of assitance, but apart from my personal
predeliction for Python, there doesnt seem to be role for it in a site
serving these kinds of loads.

Now, my predeliction for Python is extremely strong; I loathe VBscript,
and the few opportunities to program in Python I get are like holidays
for me. If anyone can suggest a role for Python in a prodcution site
serving 80 million hits a month (and growing fast), I would greatly
appreciate hearing from them.

Ive only tried Python within an obscure branch of WSH called SSH,
without success. I havent tried it as a scripting langauge replacement
for VBscript, but there has to be a significant advantage to re-
educating all the vbscript people I work with. For the work they do,
VBscript is functional and they never seem to complain about it
constraining them.

Sometime soon, this company will be developing a unix capability,
either linux or solaris. I see this as some kind of opportunity to
introduce Python programming. JPython for (prototyping?) servlets under
one of the servlet engines looks appealing. Has anyone tried extremely
high volume serving using JPython servlets?

In article <14376.24392.828292.926057 at dolphin.mojam.com>,
  skip at mojam.com (Skip Montanaro) wrote:
>
> >>>>> "Dirk" == Dirk Leas <dirkl at home.com> writes:
>
>     Dirk> Is anybody taking lots of hits with httpdapy, PyApache,
>     Dirk> ASP+Python, fastcgi, or Zope?
>
> I use a combination of Apache and ZServer (from the Zope
distribution) and a
> separate backend database server on the Musi-Cal web server:
>
>     Apache --HTTP proxy--> ZServer --XML/RPC--> db server
>
> Apache handles several virtual servers, but the main one is Musi-
Cal.  On
> behalf of Musi-Cal it handled about 65K hits in the past two days.
Many of
> those hits get proxied off to the ZServer (almost all html files are
run
> through DocumentTemplate and all search requests are passed off).  The
> ZServer handled 21K requests in the past two days.  In addition, the
> database server handles XML-RPC requests from a second machine on the
same
> physical network (www.mojam.com) that uses Mason (Apache + mod_perl +
Mason
> components).  Musi-Cal's database server handled about 160K requests
of one
> sort or another in the past two days (search URLs can result in
several
> XML-RPC calls and we do lots of other backend queries in support of
other
> activities).
>
> I figure the Musi-Cal server is handling about 1.2M web hits per
month right
> now.  It's a Pentium II 400MHz w/ 128MB of memory running RedHat
Linux 5.2
> (264 bogomips according to /proc/cpuinfo).  The load average runs
around 0.3
> during the day with bursts above 1.0 every 15-30 minutes.  The
hardware has
> been up for almost 50 days.  I restart the software servers for
various
> reasons typically once or twice a day, however.
>
> We have a more powerful dual processor machine (450+450 bogomips)
that we
> were using but set aside because of Linux SMP problems.  We've yet to
buckle
> down and address them.  Mostly we just wait for new stable revs of the
> kernel and try again.  Given the performance of our current server, I
don't
> expect to be pressed to put it back into server for awhile yet.
>
> Skip Montanaro | http://www.mojam.com/
> skip at mojam.com | http://www.musi-cal.com/
> 847-971-7098   | Python: Programming the way Guido indented...
>
>


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



From jesse at student.usyd.edu.au  Mon Nov  1 21:06:23 1999
From: jesse at student.usyd.edu.au (Jesse Sweeney)
Date: Tue, 02 Nov 1999 02:06:23 GMT
Subject: How can I make my assertions smarter?
References: <381D7087.BF8510E5@webone.com.au> <y0jpuxuryuh.fsf@vier.idi.ntnu.no>
Message-ID: <381e4350.91314552@news.usyd.edu.au>

On 01 Nov 1999 16:55:02 +0100, mlh at vier.idi.ntnu.no (Magnus L. Hetland) wrote:

[ Stuart, hungering for preconditions ]

>How about using (assuming that you won't use assert):
>
>       pre("spoon.shape == 'bent'")
>
>And then using eval() in your code? Like...
>
>  def pre(condition):
>      if not eval(condition):
>          raise "Precondition Failure", condition
>
>Pretty simple, IMO...

Simple? Yes. Works? No. <wink>

The problem is that pre's locals aren't the same as the locals in the function
from which pre is being called, so you run into problems like this one:

>>> def pre(condition):
	if not eval(condition):
		raise "Precondition Failure", condition

	
>>> def f(a, b):
	pre('a == b')
	# Some more stuff

	
>>> f(4, 4)
Traceback (innermost last):
  File "<pyshell#8>", line 1, in ?
    f(4, 4)
  File "<pyshell#6>", line 2, in f
    pre('a == b')
  File "<pyshell#2>", line 2, in pre
    if not eval(condition):
  File "<string>", line 0, in ?
NameError: a


You can fix that by passing locals() and globals() to pre(), or by using an ugly
hack like the one used to find the name of the calling function, if that's too
much typing:

import sys

def upglobals():
    try:
        1/0
    except ZeroDivisionError:
        return sys.exc_info()[2].tb_frame.f_back.f_back.f_globals

def uplocals():
    try:
        1/0
    except ZeroDivisionError:
        return sys.exc_info()[2].tb_frame.f_back.f_back.f_locals

def pre(condition, locals=None, globals=None):
    if not locals:
        locals = uplocals()
    if not globals:
        globals = upglobals()
    if not eval(condition, locals, globals):
        raise "Precondition Failure", condition


Then, we have:

>>> f(3, 4)
Traceback (innermost last):
  File "<pyshell#9>", line 1, in ?
    f(3, 4)
  File "<pyshell#8>", line 2, in f
    pre('a == b')
  File "<pyshell#6>", line 7, in pre
    raise "Precondition Failure", condition
Precondition Failure: a == b
>>> f(4, 4)
>>> 

A wiser man than me once said "It is easier to write an incorrect program than
understand a correct one." <wink>

	Cheers, Jesse.



From praschl at bigfoot.com  Wed Nov 10 20:21:31 1999
From: praschl at bigfoot.com (Michael Praschl)
Date: Thu, 11 Nov 1999 01:21:31 GMT
Subject: open( "file","mode") difference between script and commandline?
References: <941942257@bigfoot.com> <804hs1$jm2$0@216.39.151.169>
Message-ID: <MPG.129438dd349d1688989682@news.telekabel.at>

i already tried changing it to import os, and now everything works very 
well.

thanks a lot



From nbecker at fred.net  Tue Nov  9 16:44:49 1999
From: nbecker at fred.net (nbecker at fred.net)
Date: 09 Nov 1999 16:44:49 -0500
Subject: So you want graphing capability?
Message-ID: <x889047uypa.fsf@rpppc2.hns.com>

I just tried a quick experiment to allow python to talk to grace (aka
xmgrace aka xmgr aka xgr)

The latest grace-5.0.5 has a library to give access via a pipe
connection.

I wrote a simple SWIG I/F.  A little Makefile hacking to build
libgrace_np as a shared lib, and presto:

One minor problem is that GracePrintf is vararg, and swig won't work
with it.  No loss really.  Just use GraceCommand.

>python
Python 1.5.2 (#1, Sep 17 1999, 20:15:36)  [GCC egcs-2.91.66 19990314/Linux (egcs- on linux-i386
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import grace
>>> x=grace.GraceOpen(2048)
grace.GraceCommand("g0.s0 point 0, 0")
0
>>> grace.GraceCommand("g0.s0 point 1, 1")
0
>>>grace.GraceCommand("redraw")
0
>>> grace.GraceClose()
0

Now you can have REALLY nice plots.  Interested?



From ionel at psy.uva.nl  Tue Nov 16 13:00:15 1999
From: ionel at psy.uva.nl (Ionel Simionescu)
Date: Tue, 16 Nov 1999 19:00:15 +0100
Subject: wxPython and DISLIN work nicely together
References: <80s0b4$g5n@mail.psy.uva.nl>
Message-ID: <80s60j$gkq@mail.psy.uva.nl>

Ionel Simionescu <ionel at psy.uva.nl> wrote in message
news:80s0b4$g5n at mail.psy.uva.nl...
|
| DISLIN (http://www.linmpi.mpg.de/dislin/) is a powerful visualization
| package available on quite a lot of platforms, aimed at scientists.
| DISLIN is "ported" to Python as well.

Sorry,

I have to cut some from "quite a lot of platforms".
The DISLIN Python extension is only available for Win32 and Linux.

ionel






From JamesL at Lugoj.Com  Thu Nov 25 15:30:27 1999
From: JamesL at Lugoj.Com (James Logajan)
Date: Thu, 25 Nov 1999 12:30:27 -0800
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <867lj85g4c.fsf@g.local> <383CC835.B585EBDD@Lugoj.Com> <m3ogcjrn41.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <383D9C63.3A7E5806@Lugoj.Com>

Michael Hudson wrote:
> 
> James Logajan <JamesL at Lugoj.Com> writes:
> 
> > Sorry, but I think the price paid for "true" GC is too high for the
> > small number of programs that benefit from this. Perhaps one should
> > learn to clean up after oneself when cycles are expected.
> 
> While I'm inclined to agree that the loss of simplicity accompanied by
> inclusion of a true GC is too high a price to pay, "Perhaps one should
> learn to clean up after oneself when cycles are expected" is a
> ridiculous statement.

Oh yeah? Well yer mudder wears army boots! Seriously though, the price I am
talking about has absolutely nothing to do with simplicity. It has to do
with the fact that languages such as Java that have "true" GC are
non-deterministic in their execution times because GC can cause execution
pauses in the VM at unpredictable times. This makes Java unacceptable for a
large number of real-world applications that I have been involved in lately.
("I'm sorry, but your 911 call can not go through for another 20 seconds
while our Java VM garbage collects 1GB of memory on the switch.") I have
recommended Python over Java in at least one environment because of Python's
deterministic approach to memory management. Had Python not had that, I
suppose we would have gone with some other scripting language or even
designed our own (gak!).

> 
> Suppose I want to represent some chain like structure where I can
> reach any element from any other. This is not an outrageous request,
> to my mind. Clearly the easiest solution is to use some kind of doubly
> linked list, but the na?ve implementation of this will lead to a
> cyclic structure. So then either (a) you have to supply an explicit
> .finalize() method or (b) you have to hack around mightily using
> __getattr__ producing a solution that is slow and fragile. Neither of
> these is at all satisfactory to my mind. Using DC's Aquisition classes
> can help, but it's still an advanced technique that required to solve
> a problem that looks incredibly simple.

Thanks for the example of that small class of programs that *might* benefit
from "true" GC. I should point out that I have written Python programs with
cycles and it wasn't that horrid an experience making sure memory was
recovered. In my humble opinion, every experienced programmer should have
learned how to manage memory in languages like C, C++, etc. The problem with
implementing "true" GC in any language is that all known solutions (as far
as I know) make trade-offs that may be unacceptable for an important set of
applications.

(If my knowledge of "true" GC is outdated, I'd love to hear what current
solutions are!)



From Gareth.McCaughan at pobox.com  Fri Nov 26 17:40:59 1999
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: 26 Nov 1999 22:40:59 +0000
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <867lj85g4c.fsf@g.local>
	<383CC835.B585EBDD@Lugoj.Com> <86r9he2tiz.fsf@g.local>
	<383DD69E.5A927386@Lugoj.Com>
Message-ID: <867lj427wk.fsf@g.local>

James Logajan wrote:

[I said:]
>> That has the wrong semantics in the presence of side effects.
>> (It's fine for my toy example, but then so is "score<>1 and 's' or ''".
> 
> Fair enough. Your requirements weren't just "kind of like C's '?:'
> operator," they were precisely that requirement. I'm wondering though what
> great application remains unwritten because you are forced to use a
> conditional statement rather than use an operator?

None. Someone asked "what annoys you about Python". The
absence of such an operator (which, incidentally, I prefer
to think of as "like Lisp's IF", not "like C's ?:") is
something that causes me some annoyance every time I
write a non-trivial Python program.

It seems that you think I'm saying "Python is no good
because it doesn't have this". If I thought Python was
no good then I wouldn't use it and I wouldn't be reading
comp.lang.python .

And I said nothing about "requirements". It's just that
the ability to put conditionals in expressions without
ugly hacks is nice, and Python lacks it.

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
sig under construction



From markus.oberhumer at jk.uni-linz.ac.at  Tue Nov 16 00:50:00 1999
From: markus.oberhumer at jk.uni-linz.ac.at (Markus F.X.J. Oberhumer)
Date: Tue, 16 Nov 1999 06:50:00 +0100 (CET)
Subject: python-sound-server for PySol under unix
Message-ID: <XFMail.991116065000.markus.oberhumer@jk.uni-linz.ac.at>

I've written a prototype implementation of a simple
sound server for the next PySol release using a 
small Python UDP server which forks of calls
to wav players like sox and bplay, but the
performance is not suitable for general release.

Basically I'd only need what the Win32 winsound 
module provides (play/loop/stop), though further
extensions could be interesting for other applications.

So is there an official Python-way of playing 
high-quality wav sounds under Unix ? 
Using pipes, sockets, or daemons like ESD ?

If nobody has been working on this I'll hack something
together (probably a small C program using the SDL library -
http://www.devolution.com/~slouken/projects/SDL/index.html),
but this would just mean yet another non-standard sound server.

thanks,
Markus

P.S. please CC me on replies

----   Markus F.X.J. Oberhumer <markus.oberhumer at jk.uni-linz.ac.at>   ----
----             http://wildsau.idv.uni-linz.ac.at/mfx/               ----
----        5E CB 5C 85 DE AF 9E BF  E9 DA 7E 6A 39 F8 CC 67          ----

                             3 WARPS TO URANUS




From wheinemanNOwhSPAM at uconect.net.invalid  Tue Nov 16 20:20:53 1999
From: wheinemanNOwhSPAM at uconect.net.invalid (wheineman)
Date: Tue, 16 Nov 1999 17:20:53 -0800
Subject: Can't build python1.5.2 on Slackware
References: <3831ED17.5283BF03@iname.com>
Message-ID: <0113de04.b7590727@usw-ex0102-016.remarq.com>

Ok, I did this a couple months ago. In the Modules directory edit the
socketmodule.c file. Go down to the lines that read:

#elif defined(linux)
# /* whatever is currently here */
#else

and make it so:

#elif defined(linux)
#define HAVE_GETHOSTNAME_R_5_ARG
#else

and recompile.
I'm sure there some more elegant way to do this via a configure option
I just didn't have time to find it out. Hope this helps.


* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!




From robin at alldunn.com  Wed Nov 10 18:35:39 1999
From: robin at alldunn.com (Robin Dunn)
Date: Wed, 10 Nov 1999 15:35:39 -0800
Subject: IPC8: Call the 800 number!
References: <3826f6ac.0@news.bestsoftware.com>
Message-ID: <SjnW3.306$0O.5636532@news.randori.com>

Python Addict <init__self at my-deja.com> wrote in message
news:3826f6ac.0 at news.bestsoftware.com...
> There seems to be a problem between the chairs and the keyboards at the
> 703-524-6400 Key Bridge Marriott hotel -- the reservation agents don't
seem
> to know about the conference and were unable to speak to anybody that did.
> Therefore they wouldn't honor the $129 rate. I even tried rep fishing but
> got the same person again (bah!).
>
> However, the 800 number *does* know about the conference and was more than
> happy to take my money. You might want to try them first.
>

We called the 800 number and they still didn't know about it.  They did
however have an $84 room that I snached up faster than they could say "Would
you like that one?"  Now I just hope it's in the right hotel and not next to
the dumpster or something...

--
Robin Dunn
Software Craftsman
robin at AllDunn.com
http://AllDunn.com/robin/
http://AllDunn.com/wxPython/  Check it out!






From fredrik at pythonware.com  Wed Nov 10 07:09:24 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 10 Nov 1999 13:09:24 +0100
Subject: Using GNU C complier as the "default" windows complier
References: <ma4W3.7794$Ur4.223420@news.rdc1.tn.home.com>
Message-ID: <034901bf2b74$6ee42770$f29b12c2@secret.pythonware.com>

Mike Callahan <mcalla at home.com> wrote:
> This a question from a newbie but I was thinking...why doesn't the default
> Python code for windows use the GNU C complier for windows instead of the MS
> complier.

maybe because quite a few major Python contributors
happen to use MSVC?  it's a pretty good compiler, if you
happen to have it on your box...

(or maybe it's just that we're still hoping that the dist-
utils project will solve this once and for all...)

</F>




From erno at iki.fi  Sat Nov  6 16:05:33 1999
From: erno at iki.fi (Erno Kuusela)
Date: 6 Nov 1999 21:05:33 GMT
Subject: python apache module?
References: <3823D45F.CF78E6E7@mindspring.com>
Message-ID: <slrn8295t5.pr2.erno@fun77.kivikko.hoas.net>

On Sat, 06 Nov 1999 02:10:23 -0500, Rob Nikander
<nikander at mindspring.com> wrote:
>I am just beginning to learn python and so far I think it is awesome.  I
>was wondering if there is an Apache module analogous to mod_perl for
>python?  I don't see anything on the Apache page.  If not, is someone
>working on one?

there are 2.

PyApache is CGI-compatible out of the box, but is not noticeably
faster than using straight CGI because it creates and shuts down
a python interpreter for each individual http request.

httpdapy is not cgi-compatible, but it is pretty easy to use
and it is much faster then cgi/pyapache since it keeps
the python interpreter around. it is in this sense more
equivalent to mod_perl functionality.


 -- erno



From bwarsaw at python.org  Tue Nov 30 15:23:40 1999
From: bwarsaw at python.org (Barry Warsaw)
Date: Tue, 30 Nov 1999 15:23:40 -0500 (EST)
Subject: CFP Developers' Day - 8th International Python Conference
Message-ID: <14404.12876.847116.288848@anthem.cnri.reston.va.us>

Hello Python Developers!

Thursday January 27 2000, the final day of the 8th International
Python Conference is Developers' Day, where Python hackers get
together to discuss and reach agreements on the outstanding issues
facing Python.  This is also your once-a-year chance for face-to-face
interactions with Python's creator Guido van Rossum and other
experienced Python developers.

To make Developers' Day a success, we need you!  We're looking for a
few good champions to lead topic sessions.  As a champion, you will
choose a topic that fires you up and write a short position paper for
publication on the web prior to the conference.  You'll also prepare
introductory material for the topic overview session, and lead a 90
minute topic breakout group.

We've had great champions and topics in previous years, and many
features of today's Python had their start at past Developers' Days.
This is your chance to help shape the future of Python for 1.6,
2.0 and beyond.

If you are interested in becoming a topic champion, you must email me
by Wednesday December 15, 1999.  For more information, please visit
the IPC8 Developers' Day web page at

    <http://www.python.org/workshops/2000-01/devday.html>

This page has more detail on schedule, suggested topics, important
dates, etc.  To volunteer as a champion, or to ask other questions,
you can email me at bwarsaw at python.org.

-Barry



From Gaetan_Corneau at baan.com  Wed Nov 10 09:31:17 1999
From: Gaetan_Corneau at baan.com (Gaetan Corneau)
Date: Wed, 10 Nov 1999 09:31:17 -0500
Subject: Python and COM: a bug?
Message-ID: <816010E2456BD111A48700805FBBE2EEFDF796@ex-quebec-u1.baan.com>

Hello,

I have encountered a weird problem a few times in the past months. I'm
testing COM server (written here) with Python, and it seems that there is
something wrong in the code that optimizes the calls to COM objects.
Sometimes (a very rare occurence) the wrong function gets called when I
access a property. Suppose I have object O with properties a (an int), b (a
date), c (a string), then

print O.b

would give me 
"Hello, world"

The error always seem to be off by one (in the same direction), but I'm not
absolutely sure about that.

And sometimes, I get something like that:

>>> Object
<COMObject <unknown>>
>>> P = Object.Parent # Parent is an attribute that is the object that
created the current object
>>> P
<method CDispatch.Parent of CDispatch instance at c5e780>
>>> P = Object.get_Parent()
>>> P
<COMObject <unknown>>
>>>

I didn't talk about those problems 'til now because they're very hard to
reproduce and describe. And I can't just send the components to a volunteer
who would try to find the problem, because he would need to install 417 Mb
of our software to recreate the environment :(

I have workarounds, but what really annoys me is that I'm still fighting a
couple of guys here to get Python accepted: they always ask "why don't you
just use VBScript?", and it's true that I don't have those problems ith
VBScript (because VBScript doesn't try to optimize calls).

I'll try to find what the problem is by tracing the Python code, but since
I'm not familiar with it, I don't expect to find the cause. At this point,
all I want to know is if I'm the only one with that kind of strange
behavior...


Thanks,
______________________________________________________
   Gaetan Corneau
   Software Developer (Quality Assurance Team)
   BaaN  Supply Chain Solutions  
   E-mail: Gaetan_Corneau at baan.com        
   Compuserve: Gaetan_Corneau at compuserve.com  
   ICQ Number: 7395494             
   Tel: (418) 266-8252          
______________________________________________________
"Profanity is the one language all programmers know best"




From claudius at catlover.com  Tue Nov  9 16:18:28 1999
From: claudius at catlover.com (claudius at catlover.com)
Date: Tue, 09 Nov 1999 21:18:28 GMT
Subject: what is a function object?
References: <809p0a$8nf$1@nnrp1.deja.com>
Message-ID: <Ec0W3.24420$Ec1.274008@news1.frmt1.sfba.home.com>

zakons at my-deja.com says:
>Hi,
>I use function objects for sorting lists and they are great but I want
>to know more about what a function object is in the python environment.

I'll take a stab at answering this...

>* How is a function object different from other python objects? What
>does it include extra? What is missing?

>From what I gather, the most important difference is the presence of a __call__
method.

>* How may function objects be accessed by name in various namespaces?

Function objects can be referenced by name without specifying a parameter list.
I.e.

def foo(bar):
	print 'baz %s' % bar
alpha = foo
alpha('hello world') # will print out 'baz hello world'

>* Other than for sorting, what are typical uses of function objects?

Creating callbacks, storing regular functions in other datastructures?
(Excellent for writing parsers.)

MOOP (http://www.accessoft.com/moop) creates function objects that are methods
for the MOOP environment.  (And that are pickle-friendly.)  Another of my
projects (work projects, hence I can't share the source) does a similiar
thing, creating 'methods' that reside on instances (normally methods are
attributes of classes.)

>* What does it mean to say that function objects are first-class objects
>in the python environment?

Good question...I have no idea.



From jcw at equi4.com  Sun Nov  7 06:03:27 1999
From: jcw at equi4.com (Jean-Claude Wippler)
Date: Sun, 07 Nov 1999 12:03:27 +0100
Subject: ANN: MkWrap 0.9 embedded database
References: <3822C348.58034399@equi4.com>
Message-ID: <38255C7A.1791313A@equi4.com>

Eh, big oops.

Jean-Claude Wippler wrote:
> 
> This is an announcement of a new release of MkWrap for Python at:
>         http://www.equi4.com/metakit/python.html
> Ready-to-run downloads for Solaris, Linux, Windows, and Mac are at:
>         ftp://ftp.equi4.com/pub/mk/
[...]

I goofed.  They weren't "ready-to-run" at all.  They were badly broken.
Despite my "final" tests - I won't even try to explain *how* badly...
If you downloaded the 1.9d build, please download again, it's now 1.9e.

My apologies for wasting everyone's time on this.

-jcw



From Alessandro.Bottoni at think3.com  Tue Nov  2 04:25:14 1999
From: Alessandro.Bottoni at think3.com (Alessandro Bottoni)
Date: Tue, 2 Nov 1999 10:25:14 +0100 
Subject: Assorted Questions regarding the future of Python...
Message-ID: <6D8A17398E28D3119F860090274DD7DB49833A@pces.cadlab.it>

I'm evaluating Python and Visual Basic 6.0 as my possible tools for
developing a few win32 utilities for my work and I'm realizing a few
interesting facts:

1) Both these languages can be used as a "glue language" in a COM-based
application (and in a MFC-based application, too), despite the fact that
only VB is normally considered "the right one" for this task (OK: together
with Delphi).
2) Python is much clearer, smarter and better structured than Visual Basic.
So, you can make programs that are much easier to maintain.
3) Visual Basic has two advantages on Python: it can compile programs (OK: I
know that we can generate "frozen" Python programs with "Freeze") and it has
a well-organized visual IDE. 
4) As long as I can understand, Visual Basic IDE is heavily based on
pre-build "program templates". That is: you create a new application on the
basis of an existing one and use a lot of Wizards and other "visual" tools
to "fill-in the blanks". 

So, I wonder:
1) How about a Python compiler? Does anybody is working on such a project? I
saw a lot of messages regarding this topic in the past, but which is the
current status?
2) How about a Python Virtual Machine, "a la Java", as a possible
alternative to a real compiler? Is there anything like that? Is anybody
working on it?
3) How about using the "Template & Wizard" approach with Python, inside IDLE
or another IDE? Does anybody is working on such a project? Guido, do you
think to go in this direction with IDLE?
4) How about a "standard" installation tool? Is there anything like
InstallShield or "VB Package & Deployement tool" in our world? Does anybody
is working on such a project?

I will use Python for at least a part of my projects in the near future, no
matter if these tools will be available or not, but I think that a few tools
like that would make happy a lot of people and would make Python a (even
more) dreadful rival for Visual Basic.

Thanks for your attention. 
--------------------------------
Alessandro Bottoni
(Alessandro.Bottoni at Think3.com)
(abottoni at yahoo.com)
Web Programmer
Think3 inc.
(www.think3.com)




From matz at netlab.co.jp  Wed Nov  3 21:41:22 1999
From: matz at netlab.co.jp (Yukihiro Matsumoto)
Date: 04 Nov 1999 11:41:22 +0900
Subject: Comparison between Python and "Ruby"
References: <1270552684-31022246@hypernet.com>
Message-ID: <87n1svugfh.fsf@ev.netlab.co.jp>

"Gordon McMillan" <gmcm at hypernet.com> writes:

|> I don't think so.  I don't believe anyone can reduce semantic
|> variability.  Thus TMTOWTDI applies to semantics AND syntax.
|
|Did you mean to write what you wrote? From your premise, I 
|would reach the opposite conclusion. Or are you saying that 
|Guido also failed to reduce syntactic variability?

No.  I think he succeeded to reduce syntactic variability.

But TMTOWTDI (in other words: the more, the merrier) is the slogan to
emphasize maximum variability in ALL AREAS, so that I felt weird to
see Pythoneers talking TMTOWTDI or even TMTOWTHMTOWTDI in positive
attitude.  Reducing variablility in ANY AREA contradict against
TMTOWTDI, I think.

As I stated before, I don't judge whether it is good or bad.
Personally, I feel Perl went too far, though.

							matz.



From djc at itga.com.au  Wed Nov  3 00:14:01 1999
From: djc at itga.com.au (Dave Cole)
Date: 03 Nov 1999 16:14:01 +1100
Subject: A Comparison of Python and Ruby Extension Interfaces (Re: Comparison  between Python and "Ruby")
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp> <7vjtrc$hmm$1@newshost.accu.uu.nl> <38204614.2EAC6051@compaq.com>
Message-ID: <87aeowb1ie.fsf@heresy.itga.com.au>

> * The mechanism for associating methods with object types
> is somewhat simpler in Ruby. In Python, some methods live
> in slots in the type object, while others live in a method
> table or are somehow otherwise accessed through the getattr
> method. In Ruby, all methods, including the ones for built
> in operations, are attached to the object's class, which seems
> to play the role of both the Python type object and method
> table. The initialisation function of the Ruby dbm module
> starts off like this:
> 
> void
> Init_dbm()
> {
>     cDBM = rb_define_class("DBM", rb_cObject);
>     rb_include_module(cDBM, rb_mEnumerable);
> 
>     rb_define_singleton_method(cDBM, "open", fdbm_s_open, -1);
>     rb_define_singleton_method(cDBM, "new", fdbm_s_open, -1);
>     rb_define_method(cDBM, "close", fdbm_close, 0);
>     rb_define_method(cDBM, "[]", fdbm_fetch, 1);
>     rb_define_method(cDBM, "[]=", fdbm_store, 2);
>     rb_define_method(cDBM, "indexes",  fdbm_indexes, -1);
> 
> etc. There is no need to declare and initialise a type object
> or method table as in the Python version, which makes things
> somewhat more direct and easier to follow.

Couldn't you achieve the same thing in Python with an API that
allocated the PyTypeObject on the heap and returned a handle/pointer
to that.  It could also provide an internal tp_getattr which scanned
the method table you built up via method registration functions in the
same API.

- Dave



From shouldbe at message.com  Wed Nov 10 11:16:19 1999
From: shouldbe at message.com (Arinte)
Date: Wed, 10 Nov 1999 11:16:19 -0500
Subject: PyRun_SimpleString problem.
Message-ID: <80c5g9$1nm8$1@rtpnews.raleigh.ibm.com>

I am trying to run a python file from a dll.  When I run the file w/
PyRun_AnyFile it works fine.  But, when I try to run the code line by line
w/ PyRun_SimpleString I get this error on function definitions.

def bottle(n):
  File "<string>", line 1
    def bottle(n):
                ^
SyntaxError: unexpected EOF while parsing
    if n == 0: return "no more bottles of beer"
  File "<string>", line 1
    if n == 0: return "no more bottles of beer"

Can I run files line by line if they define functions?  If so, how?

Thanks





From haiyingliang1 at home.com  Fri Nov 12 19:39:04 1999
From: haiyingliang1 at home.com (Mark leung)
Date: Sat, 13 Nov 1999 00:39:04 GMT
Subject: Is it true earning while browsing?
Message-ID: <942453541.354566@news1.lynx.bc.ca>

The answer is right here.

http://www.visit.to/newsdigest



From usenet-mm-yyyy at julesallen.com  Mon Nov 15 07:27:22 1999
From: usenet-mm-yyyy at julesallen.com (Jules Allen)
Date: Mon, 15 Nov 1999 12:27:22 GMT
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <382FD44C.C85BCE7F@lemburg.com>
Message-ID: <K_SX3.58$yA4.1942@dfiatx1-snr1.gtei.net>

In article <382FD44C.C85BCE7F at lemburg.com>, M.-A. Lemburg wrote:
| Could you put a link to it from www.python.org/search ?

*Don't* do that yet! The author has a low bandwidth cap.

-- 
Want to reply via email? Put the current month and year in 
my email address.



From mlh at vier.idi.ntnu.no  Fri Nov 12 11:53:48 1999
From: mlh at vier.idi.ntnu.no (Magnus L. Hetland)
Date: 12 Nov 1999 17:53:48 +0100
Subject: Random text generator
References: <199911120349.OAA01177@envy.fulcrum.com.au>
Message-ID: <y0jk8nntzvn.fsf@vier.idi.ntnu.no>

Richard.Jones at fulcrum.com.au (Richard Jones) writes:

> [Rico Albanese]
> > Can anyone tell me if there is a random text generator for Python?  I
> > have an application that needs to generate a random password for users.
> > I have looked at the random and whrandom modules but they seem to create
> > random numbers only.
> > 
> > Thanks in advance for any advice given
> 
>    The function you want is whrandom.choice(). See
> 

Or - rather, random.choice() (which uses whrandom.choice, if
available)... No?

--

  Magnus          Echelon jamming noise:
  Lie             FBI CIA NSA Handgun Assault Bomb Drug Terrorism
  Hetland         Special Forces Delta Force AK47 Hillary Clinton 



From bennetm at kumaran.com  Tue Nov 30 06:19:53 1999
From: bennetm at kumaran.com (Bennet Rajesh)
Date: Tue, 30 Nov 1999 16:49:53 +0530
Subject: Can I develop SDI applications in Pythonwin? - A Pythonwin Newcomer
Message-ID: <821g8e$4j8$1@news.vsnl.net.in>

I understand that pythonwin is an MDI application by itself. But I am trying
to develop a SDI application in Pythonwin and have no idea where to start as
there is no proper documentation or PyCSingleDocTemplate object in
Pythonwin. Can anyone help me out.








From jimbag at kw.igs.net  Mon Nov 15 13:22:35 1999
From: jimbag at kw.igs.net (JB)
Date: Mon, 15 Nov 1999 18:22:35 GMT
Subject: exec/eval changed?
Message-ID: <942690155.686795566@news.glph1.on.wave.home.com>

Did the exec/eval statements change somehow in the 1.5 versions? I have a script
that did an eval( "some code", dict1, locals()) but now it doesn't work. I've
looked through the doc's but they are little thin as to these statements. Any
help is appreciated.

jb



From foxes at theriver.com  Tue Nov  2 14:13:43 1999
From: foxes at theriver.com (Andy Fox)
Date: Tue, 02 Nov 1999 12:13:43 -0700
Subject: How to control app from Python?
References: <38160DD8.5C695E09@theriver.com> <3817BD5F.7E05BDC9@home.com> <38185A03.9AB0FAD6@theriver.com> <7vke68$6r6$1@nnrp1.deja.com>
Message-ID: <381F37E7.6B602F33@theriver.com>

Hi Jeff,

Thanks for the reply.  Unfortunately, the app cannot be controlled via
COM.

Thanks again,

Andy Fox



From jeremy at cnri.reston.va.us  Tue Nov 16 15:19:56 1999
From: jeremy at cnri.reston.va.us (Jeremy Hylton)
Date: Tue, 16 Nov 1999 15:19:56 -0500 (EST)
Subject: Final classes like in java???
In-Reply-To: <14385.45862.55010.146256@weyr.cnri.reston.va.us>
References: <80sao7$1848$1@rtpnews.raleigh.ibm.com>
	<14385.45862.55010.146256@weyr.cnri.reston.va.us>
Message-ID: <14385.48236.202756.297241@goon.cnri.reston.va.us>

Arinte writes:
>> Is there a way you can define your python class to be final or
>> not able to be inherited from?

Fred writes:
>  Not guaranteed, but this should be good enough:
[see FinalClass below]

This doesn't really work.

>>> class FinalClass:
...     def __init__(self):
...         if self.__class__ is not FinalClass:
...             raise SomeError, "you can't subclass from me!"
... 
>>> class OhYeah(FinalClass):
...     def __init__(self):
...             pass
... 
>>> inst = OhYeah()
>>> isinstance(inst, FinalClass)

The FinalClass mechanism only makes it more inconvenient to create a
derived class.  The derived class must not call the __init__ of the
FinalClass.  If the FinalClass.__init__ doesn't do anything
significant, then it doesn't matter.  If the FinalClass.__init__ does
any non-trivial work, then the creator of the derived class needs to
do some cut-and-paste.

Jeremy





From olipt at mayo.edu  Thu Nov  4 17:51:27 1999
From: olipt at mayo.edu (Travis Oliphant)
Date: Thu, 4 Nov 1999 16:51:27 -0600
Subject: debuging in shared library (using gdb or ddd)
In-Reply-To: <3821F93B.6BF2B62A@iwr.uni-heidelberg.de>
References: <3821F93B.6BF2B62A@iwr.uni-heidelberg.de>
Message-ID: <Pine.LNX.4.10.9911041645260.9670-100000@us2.mayo.edu>

> Hi!
> 
> How can I debug a shared library module imported into python
> 
> with gdb or ddd debugger ?

I use gdb regularly to debug shared libraries imported into python:  One
way to do it is to start python, import your shared module and then in a
separate shell, run gdb and attach to the python process.  This freezes
the python processes as it is now under control of gdb.  You can then set
break points as you'd like and then enter "c" (continue) in gdb which will
start the python interpreter going again.  You can enter commands into the
interpreter and when the break point is hit it will stop and your gdb
session will be ready for commands.  This is a back-and-forth method, but
it works.

The other method that used to work for me (but doesn't anymore for some
unknown reason) is to start python from within gdb, import the module of
interest, quit python (so you are back at the gdb prompt), insert
breakpoints into the module of interest (which gdb is now aware of),
restart python and debug the shared module by running some commands in it.

As you can tell, I'm not a gdb expert, as I've typically learned just
enough to do what I want, and not learned the most elegant ways to do
things.  So, there may be better ways to do it, but the ways I've
described have worked well for me.

Travis





From jon at rdt.monash.edu.au  Mon Nov  8 17:42:46 1999
From: jon at rdt.monash.edu.au (Jonathan Giddy)
Date: 8 Nov 1999 22:42:46 GMT
Subject: __str__ vs. __repr__
References: <000001bf275c$02370d20$1d2d153f@tim> <3826A14A.3328AFAA@udel.edu>
Message-ID: <807jl6$qmf$1@towncrier.cc.monash.edu.au>

Charles Boncelet <boncelet at udel.edu> writes:

>On the dissenting side, I find the "L" useful (just as I find the
>difference between 1 and '1' useful.  Consider the following:

>>>> a,b = 2L, 2
>>>> a,b
>(2L, 2)
>>>> a**32
>4294967296L
>>>> b**32
>Traceback (innermost last):
>  File "<stdin>", line 1, in ?
>OverflowError: integer pow()

>Without the "L", it is not at all clear why a**32 works and b**32 fails.

Since the interpreter prints the __repr__ form of values, this would still 
work the same.  The proposal was to change the __str__ form of long integers.

In fact, your argument dissents against another idea in this thread that the
interpreter should display the __str__ values of results.

Personally, I tend to use the __repr__ form for details that the programmer 
needs to know about an object.  I use the __str__ form for objects that have 
a natural string conversion (exception classes are the main thing that 
spring to mind).

Someone using the interactive interpreter is not going to get very far
without 'import', and anyone who uses 'import' is, or is learning to be, a
programmer.  Therefore, I would say that interactive users should get the 
__repr__ form of the result.

Jon.



From mwh21 at cam.ac.uk  Wed Nov 10 10:29:15 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 10 Nov 1999 15:29:15 +0000
Subject: what is a function object?
References: <809p0a$8nf$1@nnrp1.deja.com> <m31z9znxy4.fsf@atrus.jesus.cam.ac.uk> <80bv3n$s9b$1@nnrp1.deja.com>
Message-ID: <m3iu3amkkz.fsf@atrus.jesus.cam.ac.uk>

zakons at eccelerate.com writes:

>   Michael Hudson <mwh21 at cam.ac.uk> wrote:
> 
> > it doesn't know a priori that the name "foo" is bound to a
> > function. So it looks the name "foo" up, and assuming it finds
> > something, attempts to call it with argument x. "Attempting to call
> > it" goes something like:
> >
> > if type(foo) is types.MethodType:
> >      apply(foo.im_func,(foo.im_self,x))
> > elif type(foo) is types.FunctionType:
> >      apply(foo,(x,))
> > elif hasattr(foo,'__call__'):
> >      apply(foo.__call__,(x,))
> 
> Thanks Michael. 

A pleasure.

> Your comment on the __call__ builtin method keyed me into Mark
> Lutz's index to Programming in Python. Buried in the way back on
> page 747, Closures in Python, he describes a class called 'counter'
> which implements __call__. What this does is make 'counter' both an
> object and a method! I believe this is what it means when folks say
> "functions are first class objects in python".

Well regular functions are objects too.

> The real question is still "OK, so why would I want to make a method
> (function) a full blown class (object)?" It seems like this makes sense
> when there is some desired side effect every time the method is called.
> (Is there a recognized pattern for this type of behavior?)

It's a way of associating state with a function. Lets have an example:
(This forms part of a snippet I wrote; see
http://tor.dhs.org/~zephyrfalcon/snippets/source/122.txt)

I was preprocessing some Python code; I wanted to replace

{{y!}}

with

y0,y1,y2,...,yn

for some n. So I used re to match this pattern:

re.sub('{{[^}]*}}',rep,line)

But what to pass in for `rep'? re.sub allows you to pass in a callback
that is passed a MatchObject describing the text to be replaced. So if
I knew n was always 2, say, I could define this:

def rep(match):
    text = m.group(0)[2:-2]
    return string.replace(text,'!','0') + "," + string.replace(text,'!','1')

but I don't; so I define

class Replacer:
    def __init__(self,n):
        self.n=n
    def replace(self,m):
        import string
        t=m.group(0)[2:-2]
        r=''
        for i in range(self.n):
            r=r+string.replace(t,'!',`i`)+','
        return r[:-1]

Then "re.sub('{{[^}]*}}',Replacer(n).replace,line)" does what I want
(NB: I could have called the `replace' method `__call__' and then
written `Replacer(n).replace' as `Replacer(n)'; that I didn't is a
personal bias of mine and makes no difference to the point).

Hmm, I wrote this code a fair old while ago so there are things I'd do
differently now and the example's a bit long - but I wanted to post an
example of when I'd found this technique actually useful.

HTH,
Michael



From xtian_the_great at hotmail.com  Mon Nov  1 20:28:51 1999
From: xtian_the_great at hotmail.com (Xtian Muirhead)
Date: Tue, 2 Nov 1999 14:28:51 +1300
Subject: dynamically generating function assignent
References: <51C9F3C392D0D111A15600A0C99E64A9AF5605@localhost> <7vlb70$95njn$1@titan.xtra.co.nz>
Message-ID: <7vlerl$a84p7$1@titan.xtra.co.nz>

Xtian Muirhead wrote...
>
<snip>
>
> Another way would be to load all the functions into a dictionary.
> e.g:
>
> def func_1():
>     <blah>
>
> def parse_func_1():
>     <blah>
>
> ...
>
> functions = {'1':(func_1,parse_func_1),
>              'a':(func_a, parse_func_a),
>                 ...}
>
> Then you could look up call1 and call2:
>
> call1, call2 = functions[valueFromFile]
> call1()
> call2()
>
> This could be easier to catch cases where the value from the file
> doesn't match any of the functions defined. (And I prefer this kind
> of static lookup to a dynamic lookup. That's just a bias, I guess.
> Suffice it to say that both of these ways are nicer than a whole heap
> of elifs. (I guess if you had hundreds of these functions, you wouldn't
> want to have to type out all of them in a dictionary (with the
> attendant likelihood of making an error).))
>
<snip>

Or maybe I'm very silly, and those functions would all be entered
automatically in globals() - then your dictionary lookup could be
from that (making sure that the resultant references did point at
callable objects).
<sigh>
(How do you cancel a usenet message when you realise you'd be
better off just lurking for a bit longer?)

Xtian

xtian at regurgitator.zzn.com

"Probably the earliest flyswatters were nothing more than some
sort of striking surface attached to the end of a long stick."






From Amos at digicool.com  Mon Nov 29 13:44:28 1999
From: Amos at digicool.com (Amos Latteier)
Date: Mon, 29 Nov 1999 13:44:28 -0500
Subject: Running apps with Zope 2?
Message-ID: <613145F79272D211914B0020AFF640193202B4@gandalf.digicool.com>

kanzow at gmd.de (Peter K.) wrote:

>I have written some web applications using the old Zope/Bobo version. I
>published them using ZopeHTTPServer, like this:
(...)
>After starting ZopeHTTPServer (on hostname.domain.de) with
>ZopeHTTPServer -p 1234 m.py
>I can access the Application object from the web:
>requesting  http://hostname.domain.de:1234/app/hello
>gets "Hello World from my Zope Object".
>This is what I want! But what must I do to achieve this in Zope 2? I
>installed it recently and can't find it in the documentation. How can I
>publish my old applications with the new Zope? Where do I have to put
>m.py? What do I have to do in the management screen? Which URL will get
>me the "Hello World..." message? Do I have to install m.py as a
Product?
>How? 

It sound like you don't want to extend Zope by writing a product; you
just want to publish your own Python objects on the web Bobo-style. 

I just wrote a small how-to which explains how to do this with a stock
Zope installation:

  http://www.zope.org/Members/Amos/ZPublisher

Let me know if this doesn't answer your question completely.

Good luck!

-Amos

--
Amos Latteier         mailto:amos at digicool.com
Digital Creations     http://www.digicool.com



From ajung at sz-sb.de  Wed Nov 10 16:11:10 1999
From: ajung at sz-sb.de (Andreas Jung)
Date: Wed, 10 Nov 1999 22:11:10 +0100
Subject: Intelligent strptime?
In-Reply-To: <19991110204023.A8220@optiplex.palga.uucp>; from gerrit.holl@pobox.com on Wed, Nov 10, 1999 at 08:40:23PM +0100
References: <19991110171256.A5196@optiplex.palga.uucp> <14377.41778.712189.7874@gargle.gargle.HOWL> <19991110204023.A8220@optiplex.palga.uucp>
Message-ID: <19991110221110.A2752@sz-sb.de>

On Wed, Nov 10, 1999 at 08:40:23PM +0100, Gerrit Holl wrote:
> It could return a tuple or something then. But if I want people to choose:
> either input "Nov, 23" or "12/6/99" or "1 December 2000", I don't want to
> check too much. There are much date processing programs (date, I think) that
> do this intelligent. You could also let the above think depend on the timezone,
Depends on your application. Usually the application (server side) sees
the environment of the server and the timezone is fixed for the server.
The standard unix "date" is also not very clever because it can not handle
e.g. german dates like 'date -d "31.12.2000"' IMHO you can not write
a one-fits-all solution without limiting your constraints.


Cheers,
Andreas






From m.faassen at vet.uu.nl  Mon Nov 15 10:46:38 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 15 Nov 1999 15:46:38 GMT
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <JlFX3.3659$s87.67930@dfiatx1-snr1.gtei.net> <80oqn9$ocq$2@newshost.accu.uu.nl> <oLUX3.286$yA4.4821@dfiatx1-snr1.gtei.net>
Message-ID: <80p9su$ppo$2@newshost.accu.uu.nl>

Jules Allen <usenet-mm-yyyy at julesallen.com> wrote:
> | Actually, 'strong leadership' is not enough. If someone cracks his whip
> | impressively in open source but doesn't write any code himself, you won't
> | get any software in the end. So you need someone who actually takes on
> | the responsibility for writing the code. :)

> Mmm! Is that the sound of you volunteering Martijn?

Definitely *not*. It is just my voice of 'experience' speaking. I don't
have enough time to volunteer for anything right now, I'm volunteering
for too many things already.

> | [Vaults of Parnassus]
> |
> | Browses through them eagerly and curiously.  :)

> The modest, retiring author

Very modest and retiring indeed. Lots of kudos to whomever it is!

> has given me a little more insight in to the
> architecture -- it's a low end machine using PyApache and he's built some
> kind of caching mechanism that's updated once a day. Mainly so the database
> doesn't get the whatsits pummeled out of it.

> It sounds like what's there is a brilliant proof of concept and with a
> little help and encouragement, it could become a robust solution to this
> problem.

I think CPAN has some automatic command-line magic system as well,
doesn't it? Is anyone familiar with it? What are the advantages of it?

Perhaps this could work well with distutils; if the Python distutils
mature enough, Parnassus can have a flag somewhere 'distutils compatible'
or something like that for each module/program in its database.

That way downloading and installing these systems
will suddenly become a lot easier. That could really make the thing
take off.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?



From wfouche at mindspring.com  Wed Nov 10 06:58:33 1999
From: wfouche at mindspring.com (Werner Fouche)
Date: Wed, 10 Nov 1999 06:58:33 -0500
Subject: Using GNU C complier as the "default" windows complier
References: <ma4W3.7794$Ur4.223420@news.rdc1.tn.home.com>
Message-ID: <80bmhf$hvl$1@nntp4.atl.mindspring.net>

"Mike Callahan" <mcalla at home.com> wrote in message
news:ma4W3.7794$Ur4.223420 at news.rdc1.tn.home.com...
> This a question from a newbie but I was thinking...why doesn't the default
> Python code for windows use the GNU C complier for windows instead of the
MS
> complier. Would it be possible to distribute the GNU package along with
the
> Python package and thus make it easier to compile, extend, and embed
Python,
> use freeze, etc? I know it would take some work, but it seems it would
make
> windows Python a lot more like unix Python. Just a thought...
>
>

I had the same idea on the weekend. So I installed Cygwin
(http://www.cygnus.com/cygwin) on my NT system and after running
"./configure" got Python to compile and run. However, I could not get Python
threads to work.





From wtopa at mindspring.com  Thu Nov 11 14:07:24 1999
From: wtopa at mindspring.com (Wayne Topa)
Date: Thu, 11 Nov 1999 14:07:24 -0500
Subject: MySQLdb Module Question
Message-ID: <19991111140723.A7271@dmcom.net>

  As a Python newbie, I am having a problem getting the
MySQLdbmodule-0.1.1 working.  I was able to use the 
MySQLmodule-0.4.1 to access my DB's but thought that I should upgrade
to the newer module.

  I can't connect to the Mysql database using
>>> import MySQLdb
>>> MySQLdb.connect('',user, passwd, Db)
Traceback (innermost last):
  File "<stdin>", line 1, in ?
	TypeError: too many arguments; expected 1, got 5

>From what I can understand from the Docs, MySQLdb.connect takes up to
9 args, so the above error really confuses me, as does this.

>>> MySQLdb.connect()
Traceback (innermost last):
File "<stdin>", line 1, in ?
File "/usr/lib/python1.5/site-packages/MySQLdb.py", line 259, in
__init__
  self.db = apply(connect, (), kwargs)

Maybe I don't have the correct Docs?

Could someone give me an example of some of the syntax of this module,
please?

TIA

Wayne
-- 
According to my calculations the problem doesn't exist.
_______________________________________________________



From wjk at wjk.mv.com  Mon Nov 15 21:56:47 1999
From: wjk at wjk.mv.com (William King)
Date: Mon, 15 Nov 1999 21:56:47 -0500
Subject: ASCII delimited files
References: <942231502.1914033961@news.interbusiness.it>
Message-ID: <3830C7EE.48A4CD2E@wjk.mv.com>


Roger Irwin wrote:

> Is there any function or module available for parsing ASCII delimited files,
> before I go and re-invent the wheel writing my own.
>
> BTW, I know it is simple to do, but I was thinking along the lines that a
> 'standard' could easily be substituted for a C version for large files.

#Was reading news and noted this discussion#on comma-delimited files so I
thought that maybe as a
#programming/python newbie I could try and contribute.

#CVT - comma delimited file input function
#read in lines of input - 1 line =='s 1 record
#pass line to function and return a list of fields
#double quotes are supposed to be in pairs (Beware cause some
#will backslash double quotes and put them in a data field) or
#data could be invalid because of a missing "
#may want to re.compile(pattern) and substitute this combo
#to something else.
#also note that it is also not uncommon to see a tab or comma
#delimited file with other character delimiters in a field
#ie.     data,"data,data",data|data|data|data
#does seem a bit 'python_awkish' -- I apologize



from string import split

def comma_Cvt( aString ):                           #function to read cvt string

     stop_N_go = 0                                        #create a guardian
     newstr = ""                                             #new string to
create
     for s_Char in aString:                           #for loop for chars in
aString
          if ( s_Char == '\"' ):                            #Look for " in
s_Char (character but really a string)
               if ( stop_N_go == 0 ):                  #Set guardian to 1 or 0
                    stop_N_go = 1
                    continue                                  #if 1st " guardian
to 1; cont.
               else:
                    stop_N_go = 0                         #if 2nd " guardian to
2; cont.
                    continue
          if ( s_Char == ','):                              #if a , as char
               if ( stop_N_go == 0 ):                     #if guardian is 0 set
the char
                    s_Char = '~~~'                             #to be ~~~  or
some other combo
                     newstr = newstr + s_Char           #build a newstring
     return split(newstr,'~~~')  #return list broken up for more


I tried this out on a couple of lines to test and see if it works -- seems too

Anyhow, just an idea -- while I learn I might as well take a stab at stuff and
sorry
havn't learned objects yet ...... Oh yeah, thanks to all who have been kind
enough to
respond to some of my NEWBIE questions.

wjk




From ivanlan at callware.com  Fri Nov 12 10:17:22 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Fri, 12 Nov 1999 08:17:22 -0700
Subject: Tkinter PhotoImage Question
References: <382C29D7.DBA3E66@callware.com> <009801bf2d1f$85953ed0$f29b12c2@secret.pythonware.com>
Message-ID: <382C2F82.7BA64FEF@callware.com>

Hi All--

Fredrik Lundh wrote:
> 
> Ivan Van Laningham <ivanlan at callware.com> wrote:
> > Ok, I've managed to create an image, using xx=PhotoImage(file), and I
> > can retrieve pixel values from that image using xx.get(x,y).
> >
> > Now I want to use the put() method to write a pixel.
> >
> 
[snip]
> # tkinter wants a list of pixel lists, where each item is
> # a tk colour specification (e.g. "#120432").  map the
> # data to a list of strings, and convert the list to the
> # appropriate tuple structure
> 
> im.put( (tuple(map(lambda v: "#%02x%02x%02x" % (v, v, v), data)),) )
> 

Oh, perfect.  That's exactly the form I've got my data in already.  I
couldn't ask for anything better.

Now I get to see how Python behaves with a list with 160,000 items in
it.  (Probably no different than a list with 1, I bet.)

> # after playing with this a little, you'll love
> # the Python Imaging Library.

I already do.  But you have to start somewhere, and it might as well be
right down there at the undifferentiated quark level;-)

<everyone-should-write-an-xlib-program.-once>-ly y'rs,
Ivan

PS:  Thanks, /F!

One of the (many) nice things about this list:  After two days, I hadn't
received any replies to my question, so I *knew* it was because the
message hadn't made it through.

This is the only mailing list I'm on for which this is invariably the
case; all the others, it's quite likely that you'll be ignored for days
or weeks on end. ...
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours
----------------------------------------------



From hniksic at iskon.hr  Fri Nov 26 09:39:00 1999
From: hniksic at iskon.hr (Hrvoje Niksic)
Date: 26 Nov 1999 15:39:00 +0100
Subject: os.execvp() weirdness
Message-ID: <9t9903l8ghn.fsf@mraz.iskon.hr>

While trussing a Python program, I found some pretty weird stuff it
was doing while trying to exec() a program.  Investigation showed that
os.execlp() called os.execvp(), but the latter, instead of being a
wrapper around the system's execvp() syscall, tries to do its own
searching.

Why is that?  All Unix systems I know of have a working execvp()
system call.  Why was it necessary to emulate one in the interpreter?
Also, os.execvp() does weird things trying to exec mktemp() file names
for reasons I couldn't discern.  None of it appears to be documented.



From davecook at home.com  Wed Nov 24 22:03:33 1999
From: davecook at home.com (David M. Cook)
Date: Thu, 25 Nov 1999 03:03:33 GMT
Subject: CRC-module
References: <38396766.B189D15A@bibsyst.no> <383C1089.1D0F7C18@bibsyst.no>
Message-ID: <slrn83p9o2.5pm.davecook@rama.escnd1.sdca.home.com>

On Wed, 24 Nov 1999 17:21:29 +0100, Thomas Weholt <thomas at bibsyst.no> wrote:

>Ok, so I`ve looked into zlib.crc32 and zlib.adler32. They seem easy
>enough to use, but I thought crc-codes had characters and numbers in
>them, 

Could the characters be a-f?  Maybe apply hex() to the number?

Dave Cook



From aa8vb at yahoo.com  Thu Nov  4 06:27:22 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Thu, 4 Nov 1999 06:27:22 -0500
Subject: Exceptions from callbacks
In-Reply-To: <3820DE5B.A5500330@lords.com>
References: <381CA8E5.5D69E64@lords.com> <7virj4$b7b$1@towncrier.cc.monash.edu.au> <381FAE6D.73F426FE@lords.com> <19991103071727.A87523@vislab.epa.gov> <3820DE5B.A5500330@lords.com>
Message-ID: <19991104062722.A184112@vislab.epa.gov>

Eric Dorland:
 |Randall Hopper wrote:
 |
 |>   (Python) -> HandlePendingMessages -> MYLIB -> Callback -> (Python)
 |
 |I understand what you mean exactly. The problem is I can't do it that
 |way. Basically, the way the library works is that you register a bunch
 |of callbacks and then call a function (eg. Loop()) that doesn't return.
 |The HandlePendingEvents (its equivalent I mean :)) isn't exposed by the
 |library, so I can't wrap it :). That's way I want a way to force the
 |interpreter to realize an exception is there. Is there a way, or a work
 |around, to do this?

I see.  So in the wrapper layer for each Python callback, you want to check
whether an exception occurred, and if so, tell Python to print it
(pretending that it has popped all the Python stack frames at that point.

So you need the C equivalent of:

      traceback.print_exc()

I don't know what this is, but hopefully someone else will chime in.

Randall





From fredrik at pythonware.com  Mon Nov 22 05:32:22 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 22 Nov 1999 11:32:22 +0100
Subject: popen2 bug?
References: <38390ACF.49D4B609@inet.polyu.edu.hk>
Message-ID: <029101bf34d4$deec6a50$f29b12c2@secret.pythonware.com>

Li Dongfeng <mavip5 at inet.polyu.edu.hk> wrote:
> After the upgrade I re-installed Python 1.5.2
> but the 
>   make test
> failed, reporting
>   test test_popen2 crashed -- exceptions.AssertionError : 

this is a known problem with the test program
in 1.5.2.  it's fixed in the CVS repository (iirc, a
patch has been posted to the newsgroup too).

> And my zope product using popen2 also crashed.

with the same error message?  spooky ;-)

(in other words, since this is a known timing
problem in the test program, maybe your pro-
duct has a similar bug?  it's not a python bug,
after all...)

</F>

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




From jimbag at kw.igs.net  Mon Nov 15 19:58:51 1999
From: jimbag at kw.igs.net (JB)
Date: Tue, 16 Nov 1999 00:58:51 GMT
Subject: exec/eval changed?
References: <942690155.686795566@news.glph1.on.wave.home.com> <m34sen35vi.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <3830AC4B.1C8C1107@kw.igs.net>

Michael Hudson wrote:
> 
> JB <jimbag at kw.igs.net> writes:
> 
> > Did the exec/eval statements change somehow in the 1.5 versions? I
> > have a script that did an eval( "some code", dict1, locals()) but
> > now it doesn't work. I've looked through the doc's but they are
> > little thin as to these statements. Any help is appreciated.
> 
> Well, eval("some code") is a syntax error. What are you trying to to
> do? We can't read minds here.
> 
> Regards,
> Michael

My apologies. I was thinking that the scoping for the 2nd and 3rd
parameters to the eval call had changed somehow and thus the code was
irrelevant to the example.

What I am trying to do is simple eval a string with python formatting
operations within and return the result. The code string is dynamic.
Basically: 

eval( "'The value of t is %(t)s' % t", self.private_dict, locals())

where t is in self.private_dict. The questions comes because this code
worked in python v1.5 but does not work in v 1.5.2 and I was curious if
there had been any changes to the scoping or to the eval/exec calls
themselves.

Sorry for the confusion.



-- 
-----------------------------------------------------------
Machine-Independent, adj.:
	Does not run on any existing machine.



From bwarsaw at cnri.reston.va.us  Tue Nov 23 12:09:17 1999
From: bwarsaw at cnri.reston.va.us (Barry A. Warsaw)
Date: Tue, 23 Nov 1999 12:09:17 -0500 (EST)
Subject: Python List
References: <m2d7t4avtb.fsf@pchoel.privat.uni-hamburg.de>
	<14393.24672.349368.650546@weyr.cnri.reston.va.us>
	<m2u2md8y3b.fsf@pchoel.privat.uni-hamburg.de>
	<14394.45433.725155.633793@weyr.cnri.reston.va.us>
	<1268755142-45000314@hypernet.com>
	<19991123071706.A1346028@vislab.epa.gov>
Message-ID: <14394.51773.666284.986665@anthem.cnri.reston.va.us>

Many people recently reported that the python-list at python.org has
seemed a bit too quiet lately, a phenomenon that can't be blamed on
the impending US Thanksgiving holiday.  I figured out what happened:
since Nov 17, nothing that was posted to comp.lang.python got gated to
the mailing list.

The good news is that I can fix this easily and that none of those
messages are lost (we never expire comp.lang.python on our news
server) so you /will/ get caught up.

The bad news is that when I unwedge this, the list is gonna get
swamped with messages.

Anyway, thanks to everyone who alerted me to the problem.

-Barry



From gerrit.holl at pobox.com  Tue Nov  2 09:45:43 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Tue, 2 Nov 1999 15:45:43 +0100
Subject: Comparison between Python and "Ruby"
In-Reply-To: <877lk1wiij.fsf@ev.netlab.co.jp>
References: <1270663667-24391282@hypernet.com> <381ED7A6.D7D7FB13@compaq.com> <877lk1wiij.fsf@ev.netlab.co.jp>
Message-ID: <19991102154543.A4469@optiplex.palga.uucp>

Yukihiro Matsumoto wrote:
>
> I guess he loves reducing variablity, if possible, which is totally
> against TMTOWTDI.  I don't judge whether it is good or bad.
> 

Are there people who see TMTOWTDI [1] as an advantage?

[1]
There's more than one way to do it.

regards,
Gerrit.
-- 
linuxgames.nl.linux.org	 All about games on Linux in Dutch (under construction).
www.nl.linux.org	       The Dutch resource for Dutch Linux documentation.
www.nl.linux.org/discoverb		    Learn foreign words and definitions.
www.nl.linux.org/~gerrit/asperger	 I'm an Asperger, go here for more info.
www.nl.linux.org/~gerrit			    About me and pictures of me.



From sonic at sega.net  Tue Nov 30 15:43:29 1999
From: sonic at sega.net (sonic at sega.net)
Date: Tue, 30 Nov 1999 20:43:29 GMT
Subject: Free classified ads
Message-ID: <RFW04.14633$sy5.21005@news20.bellglobal.com>

Costech has a free classified page on their site no gimmicks absolutely
free. Anything you wish to sell please feel free to post it there.

You may include a picture linked to your URL.


http://www.costech.com.

Thanks




From fredrik at pythonware.com  Mon Nov 29 09:54:04 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 29 Nov 1999 15:54:04 +0100
Subject: mktime() ???
References: <38427901.10B1965D@obop.com.pl>
Message-ID: <04a401bf3a79$9550af50$f29b12c2@secret.pythonware.com>

Przemys?aw G. Gawro?ski <gawron at obop.com.pl> wrote:
> In library reference it's writen, that mktime from time module raises an
> exeption when it cann't construct a correct time from a tuple parameter.
>
> I tryied the following:
>
> mktime((1999, 23, 29, 0, 0, 0, 0, 0, 0 ))
>
> (the month number is incorect),
>
> and
>
> mktime((1999, 23, 49, 0, 0, 0, 0, 0, 0 ))
>
> (the month, and day aren't correct).
>
> Well, neither one of them has raised an exeption.
>
> Is it a bug or am'I not understanding somthing in the docs ?

well, it sure looks like the C library on your computer
thinks it can construct a correct time from the given
tuple.

my Linux box doesn't mind, for example:

>>> import time
>>> t = time.mktime((1999, 23, 49, 0, 0, 0, 0, 0, 0))
>>> time.localtime(t)
(2000, 12, 19, 0, 0, 0, 1, 354, 0)

in fact, I cannot find a single platform that considers
this as an error.  it's actually documented behaviour,
at least on Solaris:

    The original values of the components may be either  greater
     than  or  less  than  the  specified  range.  For example, a
     tm_hour of -1 means 1 hour before  midnight,  tm_mday  of  0
     means  the day preceding the current month, and tm_mon of -2
     means 2 months before January of tm_year.

which seems to be exactly how most Unix and Windows
boxes behave.  maybe the library reference should be
updated?

</F>




From fdrake at acm.org  Fri Nov 12 10:15:16 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Fri, 12 Nov 1999 10:15:16 -0500 (EST)
Subject: '==' vs. 'is' behavior
In-Reply-To: <382C2C6A.64A8F2C2@theriver.com>
References: <382C2C6A.64A8F2C2@theriver.com>
Message-ID: <14380.12036.197083.409565@weyr.cnri.reston.va.us>

Andy Fox writes:
 > I thought I understood the difference between the '==' and 'is'
 > operators, but just found an exception to the rule.  Couldn't find an
 > explanation in the FAQs.
...
 > Why doesn't it work for a certain integers:
 > >>> a = 2
 > >>> b = 2
 > >>> a == b #are they equivalent?  yes.
 > 1
 > >>> a is b #are they the same object? yes.  Why?
 > 1

  It's working just fine!  Note that the "small" integers (through 99) 
are always reused, so the "is" relationship will hold.  It's not very
useful, but reusing the integer objects saves a lot of allocation of
small numbers, and makes real-world code a lot faster.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From aycock at csc.uvic.ca  Wed Nov 17 17:29:01 1999
From: aycock at csc.uvic.ca (John Aycock)
Date: 17 Nov 1999 22:29:01 GMT
Subject: ANN: (fledgling) bytecode-to-source translator
Message-ID: <80va7d$1n3k$1@uvaix7e1.comp.UVic.CA>

"Decompyle" attempts to reconstruct Python source from the bytecode in
a code object.  The latest version (pre-alpha-1) is available at

		http://csr.uvic.ca/~aycock/python

It's very fragile and is definitely far from being complete.  I'm releasing
it for demonstration and amusement purposes :-)

John



From mhammond at skippinet.com.au  Tue Nov  9 17:14:34 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Tue, 09 Nov 1999 22:14:34 GMT
Subject: Running a program as another user on Win32
References: <809ea4$ln$1@nnrp1.deja.com>
Message-ID: <e11W3.33036$pf4.31242@news-server.bigpond.net.au>

emuller at painewebber.com wrote in message <809ea4$ln$1 at nnrp1.deja.com>...
>I am writing a bunch of scripts that will need to run as Administrator
>(Domain Administrator) on Windows NT 4.0 workstation clients that
>belong to an NT domain. I've tried using LogonUser, but I always get
>the following: pywintypes.api_error: (1314, 'LogonUser', ' A required
>privilege is not held by the client.'). I've also tried using LogonUser
>with the local Administrator account on my NT machine with no luck.

You need to explicitely enable the privilege.  The following function should
do the job.  The documentation on CreateProcessAsUser is vague, and I dont
have time to test it.  It appears you need some or all of the following
privileges (all from the ntsecuritycon module): SE_ASSIGNPRIMARYTOKEN_NAME,
SE_INCREASE_QUOTA_NAME TOKEN_DUPLICATE, TOKEN_IMPERSONATE

def AdjustPrivilege(priv, enable = 1):
    # Get the process token.
    flags = TOKEN_ADJUST_PRIVILEGES | TOKEN_QUERY
    htoken = win32security.OpenProcessToken(win32api.GetCurrentProcess(),
flags)
    # Get the ID for the privilege.
    id = win32security.LookupPrivilegeValue(None, priv)
    # Now obtain the privilege for this process.
    # Create a list of the privileges to be added.
    if enable:
        newPrivileges = [(id, SE_PRIVILEGE_ENABLED)]
    else:
        newPrivileges = [(id, 0)]
    # and make the adjustment.
    win32security.AdjustTokenPrivileges(htoken, 0, newPrivileges)

Mark.





From jblazi at netsurf.de  Tue Nov 16 09:09:47 1999
From: jblazi at netsurf.de (Janos Blazi)
Date: Tue, 16 Nov 1999 15:09:47 +0100
Subject: Beginning programming with Python
References: <sng0ychgwt.fsf@motorola.com> <382B3A53.5816B2@callware.com> <382B3D79.74CD3E0F@email.sps.mot.com> <80p9ds$ppo$1@newshost.accu.uu.nl> <asteroidal-80psfo/INN-2.2.1/bat@broadway.news.is-europe.net> <38312D55.A36EB56D@iqsoft.hu>
Message-ID: <antigorite-80rol2/INN-2.2.1/attentive@broadway.news.is-europe.net>

Dear Nicholas,

what you say ist speculative. Your theses are not scientific predictions
about nature. It would be very, very difficult to design an experiment that
proves your theses wrong. Therefore from a scientific point of view, they
are neither right nor wrong. Though they contain predictions about nature,
these predictions should be much more carefully and precisely put to be
really scientific.

You are not saying that big project cannot be done with Python. You are
saying that Python is less well suited for big projects than C. But it is
not easy to atribbute a quantitative meaning to "less well". How do you want
to scale that? You could take the number of man months necessary or you
could try to quantify the quality of you product. Unless you provide us with
these details I can neither confirm nor contradict. I can only check if
there are inner contradictions in what you say. There are none, but that is
not enough. Physics as seen by Aristotle was a wonderful piece of
speculative thinking, without inner contradictions, bu today we know that we
should not overestimate its practical meaning and we better take the
Newtonian point of view.

Your second question concerns teaching.

(1)
I have not tried Python as a language for teaching yet so I have not had any
experience. I should like to use Python first and give a judgement after
using it.

(2)
I agree with you: If Python were good for nothing else, it would not be good
for teaching either. But in ther first part of your message you only say
that Python is not well suited for large projects (though, as i have tried
to show you, it is not clear, what that statement means). Now you say that
it cannot be used at all. But I have had the experience that Python is nice
for small projects, say below 3000 lines of code. So i am aifraid your
argumentation is vulnerable at this point.

Janos Blazi

Nemeth Miklos <nemeth at iqsoft.hu> schrieb in im Newsbeitrag:
38312D55.A36EB56D at iqsoft.hu...
> Janos Blazi wrote:
>
> > I have tried to use Perl in my teaching. It has a very powerful but
alas!
> > not very simple syntax! To start with, my pupils used to forget the
dollar
> > signs and Perl does not like that. So this year I shall use Python. We
shall
> > see. Time will tell, if Python is really better. Maybe my pupils will
forget
> > the indenting this time? There is actually nothing else you can forget
in
> > Python.
>
> I am quite new in Python programming, and I was a bit :-( by understanding
that
> Python does not have any (even optional) mechanism to enforce the usage of
> predeclared variables and functions. The complete lack of compile-time
> type-safety of Pythom may be a problem in large projects, which may be the
most
> serious technical (ie non-marketing) obstacle to Python's becoming a
widespread
> language like Java.
> Let us see an example:
> # Module t.py
> def f():
>   print "started"
>   x = 128
>   x.foo()
>
> Let us import it in the Python interpreter:
> >>> import t
> >>>
> You can see thet the interpreter compiled t.py to t.pyc (or t.pyo) but
found no
> error.
> All typos (typing errors of variable and function names) will be detected
ONLY
> at runtime! Python is perfect language to write programs fast, but with
the
> cost of hordes of testers. However this may not be a big problem, because
any
> piece of software should be thoroughly tested, to make sure that there are
no
> semantic errors in the application. In the case of Python the applications
> should also be tested for "syntactic" errors.
> In Python (just like in Smalltalk or in Perl, or in Tcl) there are a lot
of
> things a programmer may forget.
> I am on my way of becomming a Python programmer, but this should not
prevent me
> from seeing the pitfalls of the language.
> A tutor must emphasize these gotchas of any programming language.
> Perl has a facility called 'use strict': it is also very poor compared to
C++
> and Java strict static (ie compile time) system.
> The only really large project implemented in Python I found is Zope. Zope
is a
> fantastic piece of software, and may be regarded as an evidence of my
being
> totally wrong ;-).
>
> > I feel that it is the right language for teaching.
>
> Anyway, Python is a great language for teaching, but why to learn a
language,
> which is good only for teachning?! This is the real question you (or the
Python
> community) should answer!
>
> Miklos Nemeth
> IQSOFT
>
>
>
>





From gawron at obop.com.pl  Fri Nov 19 05:26:19 1999
From: gawron at obop.com.pl (=?iso-8859-2?Q?Przemys=B3aw?= G. =?iso-8859-2?Q?Gawro=F1ski?=)
Date: Fri, 19 Nov 1999 11:26:19 +0100
Subject: Batch window
Message-ID: <383525CB.6E2EE3AF@obop.com.pl>

Hello

Is it posible to automaticaly minimize the python window, when starting
a GUI application on Windows ?

Thankx

Przemek
--
Przemyslaw G. Gawronski UIN:8358522
mailto:gawronskip at usa.net
mailto:gawron at obop.com.pl





From wtanksle at hawking.armored.net  Thu Nov  4 16:07:33 1999
From: wtanksle at hawking.armored.net (William Tanksley)
Date: 4 Nov 1999 21:07:33 GMT
Subject: ~ 1 = -2 ???
References: <m2vh7hk1ys.fsf@netmeg.net>
Message-ID: <slrn824m9l.ngk.wtanksle@hawking.armored.net>

On Fri, 05 Nov 1999 04:01:22 GMT, Les Schaffer wrote:
>i am doing some bit operations, and dont understand the following:

>>>> ~ 1
>-2

>huh???? i would have thought that bitwise invert of 1 would be
>0001 --> 1110 = 13 (base 10)
>what marble am i missing?

The marble about 32-bit twos' complement signed integers.  They're not
four-bit, BTW.

To find the negative in twos' complement, which Python uses, you NOT the
bits and add one.  So if you only NOT the bits, you'll get one less than
the negative of the number you started with.  Try ~-5 for an example.

The bit pattern is still what you need if you're worried about bit
patterns, though.  Personally, ALL of my coding uses unsigned numbers.

>les schaffer

-- 
-William "Billy" Tanksley



From alan.gauld at gssec.bt.co.uk  Tue Nov  9 11:02:54 1999
From: alan.gauld at gssec.bt.co.uk (Alan Gauld)
Date: Tue, 09 Nov 1999 16:02:54 +0000
Subject: Learning to program
References: <7v82o9$c31$1@news4.svr.pol.co.uk>
Message-ID: <382845AE.9023D50F@gssec.bt.co.uk>

Jonathan wrote:
> A friend of mine wants to learn to program. Python seems one of the most
> suitable languages to start with. I think the Lutz et al book is good at
> explaining the language to a newcomer, but I'd also like to find an
> accessible book that explains program design, clarity, debugging, data
> structures, etc in a Python friendly way. 

Thats pretty much what I try to do in my online tutor...

http://members.xoom.com/alan_gauld/tutor/tutindex.htm

Any feedback is very welcome.

Alan G.

-- 
=================================================
This post represents the views of the author 
and does not necessarily accurately represent 
the views of BT.



From mikael at isy.liu.se  Tue Nov  2 07:43:22 1999
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Tue, 02 Nov 1999 14:43:22 +0200 (MET)
Subject: commas vs. spaces. vs. the legions of new lines
In-Reply-To: <jwbaxter-3110992055260001@otter.olympus.net>
Message-ID: <XFMail.991102144322.mikael@isy.liu.se>

On 01-Nov-99 John W. Baxter wrote:
 >  Hmmm...don't say all those bad things about spaces in France (where one
 >  would write 12 345,67 in the place a US person would write 12,345.67). 
 >  Or Sweden or Norway for Finland.  And then there is the Spanish or Flemish
 >  12.346,67 (which doesn't have spaces, but doesn't use , and . in the
 >  "right" places).  Sigh.

This problem is actually a PITA in an increasingly international world.
FWIW, you may encounter 12 345,67 aswell as 12.346,67 and sometimes
12 345.67 in Sweden (but never 12,345.67). My UNIX box at work prefers
12 345.67, while my Winows box at home prefers 12 345,67. Perhaps I can 
change that, but I'm still too lazy to do anything about it.

Actually, it is ridiculous that such a thing has to be a problem, 
especially since there is an international standard out there, as was
pointed out by Russel W Smith.

standards-are-great--everyone-should-have-one-ly y'rs

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    02-Nov-99
Time:    14:20:53

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



From philh at vision25.demon.co.uk  Thu Nov 25 15:47:52 1999
From: philh at vision25.demon.co.uk (Phil Hunt)
Date: Thu, 25 Nov 99 20:47:52 GMT
Subject: Python complaints
References: <XFMail.991124113221.mikael@isy.liu.se> <v04003a01b46194dec84b@[192.168.24.43]>
Message-ID: <943562872snz@vision25.demon.co.uk>

In article <v04003a01b46194dec84b@[192.168.24.43]>
           rgacote at AppropriateSolutions.com "Raymond Cote" writes:
> Thomas Hamelryck <> wrote:
> > I don't want to start the discussion about indentation again; it's no use
> > since indenting blocks is here to stay. The point is that if you want to 
> > write code that looks ugly or unreadable you will succeed in doing that,
> > indentation or not. In many cases, using a variable indentation (depending
> > on what construct you are using or the length of a line) leads to code
> > that is visuably much more appealing.
> 
> I was sure this 'indenting thing' would be one of the two major issues that
>  kept me grumbling about Python. However, after just a month or so of casual
>  Python programming, I find myself becoming dissatisfied with extra work
>  required by braces and begin/end in my day to day C and Pascal programming.
> 
> Was totally surprised how quickly I came to like the indenting.  

My experience was very similar.


-- 
Phil Hunt - - - phil at comuno.com
Want to see the FAQs for your favourite newsgroups? Read them on
the web with Comuno's FAQ server at http://www.comuno.com/faq/ 




From quinn at groat.ugcs.caltech.edu  Mon Nov 29 16:28:48 1999
From: quinn at groat.ugcs.caltech.edu (Quinn Dunkan)
Date: 29 Nov 1999 21:28:48 GMT
Subject: Unbuffered stdout
References: <51C9F3C392D0D111A15600A0C99E64A9AF568E@localhost>
Message-ID: <slrn845s25.pja.quinn@groat.ugcs.caltech.edu>

On Mon, 29 Nov 1999 09:47:22 -0700, Bjorn Pettersen <bjorn at roguewave.com> wrote:
>> leon <leonar_n at epita.fr> wrote:
>> > I'm developping an IRC like software which must work on the 
>> Web. I can't use
>> > Java, and I'm using Python to program the CGI. But, I have 
>> a big problem : I
>> > don't know how to make the stdout unbuffered.
>> > I tried to make some
>> >     sys.stdout.write("test")
>> >     sys.stdout.flush()
>> > but it don't work.
>> 
>> looks fine to me.  maybe something else
>> is wrong?
>> 
>> anyway, you could try running the inter-
>> preter with the -u option:
>> 
>> $ python -?
>> Unknown option: -?
>> usage: python [option] ... [-c cmd | file | -] [arg] ...
>> Options and arguments (and corresponding environment variables):
>> ...
>> -u     : unbuffered binary stdout and stderr (also PYTHONUNBUFFERED=x)
>> ...
>> 
>> </F>
>
>Note that some webservers automatically buffer cgi output until the cgi
>closes stdout.  To the best of my knowledge there are no pretty workarounds
>for this...  You should check your webserver's docs to see if this is the
>case though.

apache 1.3 and on has 'effectively unbuffered' cgi scripts.  For apache 1.2
and earlier you can use the nph- hack.



From thor at localhost.localdomain  Mon Nov  8 12:15:54 1999
From: thor at localhost.localdomain (Manuel Gutierrez Algaba)
Date: 8 Nov 1999 17:15:54 GMT
Subject: pythonistas or pythoneers?
References: <Pine.LNX.4.10.9911081026470.4450-100000@robert.polytopic.com>
Message-ID: <slrn82e4mp.43f.thor@localhost.localdomain>

On Mon, 8 Nov 1999 10:29:01 -0600 (EST), Robert Meegan <rmeegan at murl.com> wrote:
>On Sun, 7 Nov 1999, Juergen A. Erhard wrote:
>
>> 
>> I'm confused: are we pythoneers or pythonistas?

I'm quite amazed with these new word : "pythonistas"

Because it sounds too much Spanish. You see, here ( Spain), 
the football team supporters usually bear the suffix -istas. 
So we have Madridistas ( Real Madrid), Barcelonistas ( FC Barcelona)
or Sevillistas ( FC Sevilla).

So pythonistas sounds extremely powerfull, because python can 
be spelled in Spanish as "piton", that means "horn" ( like the bull's
horns). So pythonistas sounds like the "damned crazy followers 
of the horns" :DD
Certainly it's a worth Slogan for all pythoneers, that sounds more
the way as bucaneer, thus regarding a bit the Dutch origin of 
python. So it's a good Slogan ( cry for war) too.

>> 
>> I tend to go for pythoneers... sounds more pythonic.
>> 
>> Or are these two levels of experience?  If so, what's higher?
>> 
>> Confused,
>> 
>
>I always thought we were Pythogorians.
>
>
>Your most righteous servant,
>
>--- Robert
>
>
>Robert Meegan
>
>Murl.com - 
>Because it's not just your data, it's your life.
>
>
>


-- 




From pinard at iro.umontreal.ca  Mon Nov  1 10:46:17 1999
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 01 Nov 1999 10:46:17 -0500
Subject: commas vs. spaces. vs. the legions of new lines
In-Reply-To: "G. Lewis"'s message of "Mon, 01 Nov 1999 03:58:53 GMT"
References: <7vj35t$8p1$1@nnrp1.deja.com>
Message-ID: <oqiu3mi59y.fsf@titan.progiciels-bpi.ca>

"G. Lewis" <lemiss at my-deja.com> writes:

> c.l.py has an exellent reputation for being extremely friendly and
> helpful---i'm astonished at the general endless patience I've seen
> exhibited here, even for the most inane and pathetic newbie FAQs.
> Lets keep it!

By all means, I agree.  I moved to Python from Perl mainly _because_ of the
difference in people attitudes, and only learned to really appreciate the
language after the fact.  My only fear is that Python will grow popular
enough so the sad history of c.l.py repeats here, so I hope that Python
will spread slowly enough so the spirit gets preserved! :-)

> So the unix old-timers <wink> know a billion and one ways to convert Dos
> to Unix text files.  Why discourage someone from making a pure python
> implementation... and a decent attempt at it as well, with subdirectories
> and attempts at some sort of "smart" file type guessing.  Not a tool I'd
> use, myself, but surely it has value.  Or even if it didn't, still surely
> it does not deserve denegration.  Not one person complimented his code.

I was one of those who replied to this thread, and feel concerned by your
remarks.  Yes, I should have acknowledged the effort, at least because I
understand what it means, and am sorry for not having done so.  On the other
hand, to compliment some code, one has to take a look at it, and not being
Internet-connected usually (I use UUCP), it requires a special effort on me
which I often try to spare.  If someone saw denegration in what I said, then
I was surely misinterpreted -- as I did not have such feelings, really not.

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





From T.Moddleton at news.vex.net  Wed Nov 17 04:14:35 1999
From: T.Moddleton at news.vex.net (T. Moddleton)
Date: 17 Nov 1999 09:14:35 GMT
Subject: FAQ
References: <80toj9$6rf$1@ellebore.saclay.cea.fr>
Message-ID: <80trlr$2utq$1@hub.org>

On Wed, 17 Nov 1999 09:23:36 +0100, patrick Hede <patrick.Hede at cea.fr> wrote:
>Hi, where is the FAQ?

http://www.python.org

However due to a power outage the site is down for a few more hours.

--snip--
CNRI's building will
be out between 2:00am and 6:00am Wednesday November 17th, EST. 




From none at fake.net  Fri Nov 12 00:17:19 1999
From: none at fake.net (me)
Date: Fri, 12 Nov 1999 05:17:19 GMT
Subject: Win32api.... can it create a Windows shortcut (.lnk) file?
References: <137053459BA7D211B3930090272A62CA8EDD80@newsintern.dspace.de>
Message-ID: <zpNW3.53035$Rx2.650681@news1.rdc1.bc.home.com>

That's a nice clean example.... but I'm having one small problem, even
taking it exactly as you put it:


##########################################
Traceback (innermost last):
  File "C:\Program
Files\Python\Win32\Pythonwin\pywin\framework\scriptutils.py", line 237, in
RunScript
    exec codeObject in __main__.__dict__
  File "C:\PROGRA~1\PYTHON\MYPROJ~1\WINTOOLS\wintools.py", line 144, in ?
    CreateShortCut(Path,Target,Arguments,StartIn,Icon,Description)
  File "C:\PROGRA~1\PYTHON\MYPROJ~1\WINTOOLS\wintools.py", line 132, in
CreateShortCut
    persist.Save(Path, 1)
com_error: (-2147467259, 'Unspecified error', None, None)
##########################################

Unfortunately, "Unspecified error" doesn't help me too much... q:].  I'll be
looking into that tonite, but if you have a quick fix, I'd appreciate it.
I've been using the win32api module a lot, and it's great, but not pythoncom
a lot yet...

q:]

Kevin.


Stefan Migowsky <smigowsky at dspace.de> wrote in message
news:137053459BA7D211B3930090272A62CA8EDD80 at newsintern.dspace.de...
> Hi
>
>
> Mark Hammond did a great job on wrapping the win32api. So
> creating a shortcut is just so easy :
> ################ CreateShortCut.py ####################
>
> from win32com.shell import shell
> import win32api
> import pythoncom
> import os
> import sys
>
> def CreateShortCut(Path, Target,Arguments = "", StartIn = "", Icon =
("",0),
> Description = ""):
>     # Get the shell interface.
>     sh = pythoncom.CoCreateInstance(shell.CLSID_ShellLink, None, \
>         pythoncom.CLSCTX_INPROC_SERVER, shell.IID_IShellLink)
>
>     # Get an IPersist interface
>     persist = sh.QueryInterface(pythoncom.IID_IPersistFile)
>
>     # Set the data
>     sh.SetPath(Target)
>     sh.SetDescription(Description)
>     sh.SetArguments(Arguments)
>     sh.SetWorkingDirectory(StartIn)
>     sh.SetIconLocation(Icon[0],Icon[1])
>
>     # Save the link itself.
>     persist.Save(Path, 1)
>
> if __name__ == "__main__":
> TempDir = os.environ["TEMP"]
> WinRoot = os.environ["windir"]
>
> Path        =  WinRoot    + "\\Profiles\\All Users\\Desktop\\New
> Link.lnk"
> Target      =  Pythonroot + "pythonw.exe "
> Arguments   =  TempDir + "\\test.py"
> StartIn     =  TempDir
> Icon        = (Pythonroot + "\\py.ico", 0)
> Description = "New Link"
>
> CreateShortCut(Path,Target,Arguments,StartIn,Icon,Description)
>
>
> Stefan
>
>
>
> Dipl.-Physiker Stefan Migowsky
> dSPACE GmbH
> Technologiepark 25 33100 Paderborn (Germany)
> mailto:smigowsky at dspace.de http://www.dspace.de
>
>
>
> -----Original Message-----
> From: me [mailto:none at fake.net]
> Sent: Thursday, November 11, 1999 7:54 AM
> To: python-list at python.org
> Subject: Win32api.... can it create a Windows shortcut (.lnk) file?
>
>
> I'm just trying to find an easy to create shortcuts on Windows platforms
> through Python... it doesn't seem to be a simple matter though.
>  I've gone
> through all the Win32 docs, and found nothing, and have started writing
> (with limited success) a C function to do it.
>
> Is there an easy way?  q:]
>
> Thanks,
> Kevin Cazabon
> kcazabon at home.com
>
>
>
> --
> http://www.python.org/mailman/listinfo/python-list
>
>





From vykron at my-deja.com  Tue Nov 23 20:32:08 1999
From: vykron at my-deja.com (vykron at my-deja.com)
Date: Wed, 24 Nov 1999 01:32:08 GMT
Subject: Block and continue question
Message-ID: <81ff6o$rrg$1@nnrp1.deja.com>

I've embedded Python within my c++ app.  I've
been looking through the
Python API and am unsure of the easiest way to
implement block and continue.
I would like to write functions of the form:

somemodule.WaitForEvent();

WaitForEvent() would call a relevent C function
to register the waited for
event.  When the script reaches the above line
the interpreter would exit
and my main application would continue as normal
until the specified event
occured.  When the event occurs I would like my C
code to reenter the Python
interpreter at the appropriate line.

What is the most elegant way to do this within
Python?

thanx.
Kelvin McDowell
kelvinm at relic.spam.spam.spam.chips.com


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



From mwh21 at cam.ac.uk  Mon Nov  8 18:00:35 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 08 Nov 1999 23:00:35 +0000
Subject: IDLE and rlcompleter [Was: Re: making the Python case to my dept.]
References: <9911051740130D.02530@william.jinsa.org> <3823FC43.CF7FBD71@python.net> <m3puxno6g0.fsf@atrus.jesus.cam.ac.uk> <382523fc.14124721@news.usyd.edu.au>
Message-ID: <m366zcoagc.fsf@atrus.jesus.cam.ac.uk>

jesse at student.usyd.edu.au (Jesse Sweeney) writes:

> On 06 Nov 1999 17:50:23 +0000, Michael Hudson <mwh21 at cam.ac.uk> wrote:
> 
> [ snip ]
> 
> >``import rlcompleter''!!
> >
> >It makes life *so* much easier, and is the reason I still use the
> >basic interpreter rather than idle.
> >
> >Hmm, implementing completion in idle? It has ``expand word'' but
> >that's not what I want. Would this be easy? Pointers appreciated.
> 
> Coincidentally, I had the same idea just a few days ago, and after a
> bit of tinkering, I came up with two different IDLE extensions. I've
> put the files on the web at
> 
> http://www-personal.usyd.edu.au/~jsweeney/python .
> 
> The first (RLComplete.py) attempts to mimic the behaviour of
> readline and rlcompleter in the Python shell, and the second
> (AutoComplete.py) is a bastard child of rlcompleter and the current
> "Expand Word".

Funky!

I had to change <Meta-.> to <Meta-period> and then to <Alt-period>
(the latter is probably a misconiguration at this end, but the first
may not be).

I'm very used to some other readline conventions such as Meta-. being
yank-last-arg and C-r being reverse-search-history and some
others. Would these be easy to add to your extension? (I know next to
nothing about Tkinter).

> Obviously, unlike "Expand Word" these only work in the Python shell,
> not in the file-editing windows. (Now that I think about it, this
> would probably be pretty useful in those editing windows, but that
> would take some partial compilation of some sort I suppose, which is
> way too much work for me.)

I think I'd settle for the completions being selected from the words
known to the Python shell, if that helps.

While we're on the subject of idle, is there any easy (preferably one
or two keystrokes) that way to import/reload the module being edited
into the Python shell and bring said shell to the front for instant
interactive play? That'd be ?bercool.

> To add these to IDLE, put the two files in your IDLE directory, (or
> just the one you want -- I'm flexible!), and add "RLComplete" and
> "AutoComplete" to the list of names in extend.py (also in your IDLE
> directory).
> 
> Problems and inconsistencies are documented in the module
> docstrings. Let me know of any bugs.
> 
> 	Cheers, Jesse.

Good work!

Cheers,
Michael



From fredrik at pythonware.com  Wed Nov 10 08:04:02 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 10 Nov 1999 14:04:02 +0100
Subject: SWIG and function name question
References: <7vmc7e$8h$1@marcie.fissiontech.com> <m3ln8f9y9u.fsf@atrus.jesus.cam.ac.uk> <7vqojf$bgf$1@marcie.fissiontech.com>
Message-ID: <002201bf2b7c$106f6c60$f29b12c2@secret.pythonware.com>

Scott Barron <kain at twilight.> wrote:
> >>> import bar
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
>   File "bar.py", line 3, in ?
>     class Foo:
>   File "bar.py", line 4, in Foo
>     self.f = foo.__foo
> AttributeError: _Foo__foo
> 
> If I comment out self.f = foo.__foo all works fine.  What am I doing wrong
> here?

names that start with "__" are private to the class
they're defined in.  foo.__foo is interpreted as a
private member of the Foo class, not the global
variable.

</F>

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





From jima at aspectdv.com  Wed Nov  3 12:39:07 1999
From: jima at aspectdv.com (Jim Althoff)
Date: Wed, 03 Nov 1999 09:39:07 -0800
Subject: Comparison between Python and "Ruby"
In-Reply-To: <slrn81ve6m.l1c.neelk@brick.cswv.com>
References: <1270663667-24391282@hypernet.com>
 <381ED7A6.D7D7FB13@compaq.com>
 <877lk1wiij.fsf@ev.netlab.co.jp>
 <19991102154543.A4469@optiplex.palga.uucp>
 <381F9EB4.63FA033F@python.net>
Message-ID: <199911031734.JAA15489@hermes.aspectdv.com>

At 03:42 AM 11/3/99 +0000, Neel Krishnaswami wrote:
. . .
>
>The differences are that there's less of a closed-world attitude in
>Python than in the ST world, though -- Python can be used for sysadmin
>work where ST (even Squeak) wouldn't be a natural fit. But on the flip
>side Smalltalk has always had a stronger focus on UI and dev
>environments.  This is changing, though -- the ST culture is beginning
>to value the ability to deal with the world outside the system image,
>and Guido's latest enthusiasm is "computer programming for everybody"
>(which is a very ST-ish sort of idea).

Absolutely right.  Part of the original dream of the Smalltalk development
team was to build a programming language/environment that would be so
easy that anybody -- even kids -- could use it.  Hence the name "Smalltalk".

Jim




From mcfletch at vrtelecom.com  Mon Nov 15 22:13:40 1999
From: mcfletch at vrtelecom.com (Mike Fletcher)
Date: Mon, 15 Nov 1999 22:13:40 -0500
Subject: list.without()?
In-Reply-To: <y0jln7zi9pb.fsf@vier.idi.ntnu.no>
Message-ID: <NDBBIAPHEKGOBIHCBLIICEOMCAAA.mcfletch@vrtelecom.com>

Well...

def without( source, element):
	temp = source[:]
	while temp:
		try:
			temp.remove( element )
		except:
			break
	return temp

Seems to work.  I'm not sure if by "standard language" you were saying "in
the standard libraries", but there's certainly no non-standard Python in the
above.

Cheers,
Mike

-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Magnus L. Hetland
Sent: November 15, 1999 7:04 PM
To: python-list at python.org
Subject: list.without()?



Hi!

Is there a function or method in the standard language that can
non-destructively return a list without a specified element? Would it
be possible to add it as a method to lists?

i.e:

# pseudocode

class list:
    def without(self,element):
        temp = self[:]
        temp.remove(element)
        return temp

It might be useful - and sure beats using a temp-variable ;)

--

  Magnus          Echelon jamming noise:
  Lie             FBI CIA NSA Handgun Assault Bomb Drug Terrorism
  Hetland         Special Forces Delta Force AK47 Hillary Clinton

--
http://www.python.org/mailman/listinfo/python-list




From vroonhof at frege.math.ethz.ch  Fri Nov  5 21:55:34 1999
From: vroonhof at frege.math.ethz.ch (Jan Vroonhof)
Date: 06 Nov 1999 03:55:34 +0100
Subject: A newbie dictionary question
Message-ID: <byr9i42us9.fsf@bolzano.math.ethz.ch>

I have the distinct feeling this is FAQ, but if so I missed it. It is
late (or better early) here.
I am trying to teach  myself python with a little problem I have that
involves parsing ChangeLog files.

Thus I wrote a module that creates a dictionary-like class with the
text of the ChangeLog entry as the key. Later I want to use this to
check whether specific changes were applied. However I want to do the
comparisons regardless of whitespace. Thus I a made small class that
has __cpm__ and __hash__ functions that ignore whitespace.

However looking up objects in the hashes doesn't seem to work. Can
somebody tell me what obvious thing I am missing?

The code is below. Note that I added the complete module (which is 20
lines or so more than a trivial example) in the hope that somebody
comment on it stylewise as this is my first Python program.
In particular I have the feeling the while loops in the scan(self)
function could be done much clearer

#!/usr/bin/env python
"""This class implements a dictonary like class,
ChangeLog, that represents entries from ChangeLog file"""
import re
import sys
import string
from UserDict import UserDict

authpat = re.compile(
    r"^(\d\d\d\d-\d\d-\d\d|\w\w\w \w\w\w\s+\d+\s+\d\d:\d\d:\d\d \d\d\d\d)"
    +r"\s+([^<]+)\s+<(.+)>\s?$")
emptypat = re.compile(r"^\s*$")
startpat = re.compile("^(?:\t| {4})\\s*\\* (.*)")
contpat =  re.compile("^\\s*(?:\t| {4})\\s*(\\S.+)")

_debug = 0

class ChangeAuthor:
    def __init__(self,name,address):
        self.name = name
        self.address = address

    def __str__(self):
        return self.name + "<" + self.address + ">"
    
class ChangeEntry:
    def __init__(self,value):
        self.value = value
    
    def __str__(self):
        self.value

    # bad approximation?
    def nospaces(self):
        return string.join(string.split(self.value))

    def __cmp__(self,other):
        return self.nospaces() == other.nospaces()

    def __hash__(self):
        return hash(self.nospaces())
    

class ChangeLog(UserDict):        
    def scan(self,fd):
        author = None
        while 1:
            line = fd.readline()
            if author:
                if not line: break
                if emptypat.match(line): continue 
                while 1:
                    s = startpat.match(line)
                    if not s:
                        break
                    entry = s.group(1)
                    while 1:
                        line = fd.readline()
                        if not line: break
                        if emptypat.match(line): continue
                        s = contpat.match(line)
                        if not s: break
                        entry = entry + "\n"+ s.group(1)
                    if _debug: print entry
                    self[ChangeEntry(entry)] = author
                if _debug: print author
            if not line: return
            m = authpat.match(line)
            if m: author = ChangeAuthor(m.group(2),m.group(3))

    def scanfiles(self,files):
        for file in files:
            fd = open(file)
            self.scan(fd)
            fd.close()
        
    
            
def _test():
    global _debug
    _debug = 0
    args = sys.argv[1:]
    if not len(args):
        args = [ "ChangeLog"]
    clonce = ChangeLog()
    clonce.scanfiles(args);
    _debug = 0
    cltwice = ChangeLog()
    cltwice.scanfiles(args)

    # Verify  that the keys are actually equal and have the same has value
    # I am betting that for dictonaries constructed in the same way the order
    # of keys() is identical.
    for n in range(0,len(clonce.keys())):
        keyonce = clonce.keys()[n]
        keytwice = clonce.keys()[n]
        if keyonce != keytwice:
            raise RuntimeError, "Keys Different"
        if hash(keyonce) != hash(keytwice):
            raise RuntimeError, "Hash Different"

    for v in cltwice.keys():
        cltwice[v]
        clonce[v]   # This gives a KeyError, why?
         

if __name__ == '__main__':
	_test()




From doughellmann at home.com  Wed Nov  3 07:05:18 1999
From: doughellmann at home.com (Doug Hellmann)
Date: Wed, 03 Nov 1999 12:05:18 GMT
Subject: Drawing (directed) graphs
References: <ye1bt9das2p.fsf@ex.ac.uk>
Message-ID: <3820264D.4ADAF3C0@home.com>

You might want to check out daVinci: http://www.informatik.uni-bremen.de/~davinci

The purpose of the app is to do very nearly what you wanted -- visualize
digraphs.  It handles optimizing the layout to reduce edge crossings and
minimize the size of the overall graph.  I have feed it hundreds of
nodes representing modules within some utility libraries my company
developed, and it happily laid out the graph for me so I could find ways
to reduce interdependencies in the code.

The last time I used it the easiest way to add/remove nodes was to edit
the data file, but now there is a new API that can be used.  I have a
class to create the graph files from Python code, if you would find that
useful I can send it to you.  I also have some code somewhere that
processed Makefile like input files (easier to construct than a daVinci
input file) and created graph files to be visualized.

Doug

"R.M.Everson" wrote:
> 
> Hello,
> 
> Does anyone know of Python code for drawing graphs of the
> graph-theoretic type? Ideally, I'd like to be able to interactively
> add nodes and edges etc and add data to structures associated with the
> nodes and edges.
> 
> Many thanks
> Richard Everson.
> 
> --
> Exeter University     R.M.Everson at ex.ac.uk
> http://www.dcs.ex.ac.uk/academics/reverson



From parkw at better.net  Thu Nov 11 17:05:57 1999
From: parkw at better.net (William Park)
Date: Thu, 11 Nov 1999 17:05:57 -0500
Subject: Beginning programming with Python
In-Reply-To: <sng0ychgwt.fsf@motorola.com>
References: <sng0ychgwt.fsf@motorola.com>
Message-ID: <19991111170557.A2523@better.net>

On Thu, Nov 11, 1999 at 02:12:02PM -0700, Aaron Ginn wrote:
> 
> Hello all,
> 
> I'm starting a non-profit organization dedicated to promoting computer
> literacy in low-income youth (ages 12-18).  The curriculum will
> include an introduction to programming using either Python or Perl.
> 
> First, I know Perl very well, but I know nothing about Python.  I've
> heard from many sources that Python is the best language to teach
> someone with no programming experience because it emphasizes
> fundamentals and the correct (i.e. best) way to program.  Could anyone
> comment on how the two languages compare in this aspect?  Also, I will
> need to learn Python as well if I choose to go that route.  I am very
> literate with Perl and Tcl/TK.  How quickly can I expect to pick up
> Python?

One weekend... at least for me :).  Go through the tutorial and other
documentations in <www.python.org/doc/>.

> 
> Secondly, if I choose to use Python for this, I would like to supply
> each student with a reference book on Python.  I've narrowed it down
> to a choice between the two O'Reilly books, "Programming Python" and
> "Learning Python".  PP apears to have a lot more meat to it, but LP
> appears to be more suited to the beginner.  Assuming the audience
> includes kids that have never programmed before, should I go with LP,
> or will I miss out on anything by not getting the more detailed book?

For beginners, "Learning Python" is definitely better than "Programming
Python".  But, as announced in <www.python.org>, ToExcel is publishing
the official tutorial, reference, and library documentations; so, you
may want to look into those.  Also, there are other books coming out
now.

William



From a.eyre at optichrome.com  Tue Nov  2 09:44:17 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Tue, 2 Nov 1999 14:44:17 -0000
Subject: Drawing (directed) graphs
In-Reply-To: <ye1bt9das2p.fsf@ex.ac.uk>
Message-ID: <001701bf2540$bd4d07e0$3acbd9c2@peridot.optichrome.com>

> Does anyone know of Python code for drawing graphs of the
> graph-theoretic type? Ideally, I'd like to be able to interactively
> add nodes and edges etc and add data to structures associated with the
> nodes and edges.

Depends if you're looking for a proper graphics display, or an ASCII
representation. A full graphics version would not be platform
independent, and would require a graphics library.

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From jima at aspectdv.com  Mon Nov 15 16:35:09 1999
From: jima at aspectdv.com (Jim Althoff)
Date: Mon, 15 Nov 1999 13:35:09 -0800
Subject: Comparison between Python and "Ruby"
In-Reply-To: <80it3h$3d2$1@green.twinsun.com>
References: <1270663667-24391282@hypernet.com>
 <381ED7A6.D7D7FB13@compaq.com>
 <877lk1wiij.fsf@ev.netlab.co.jp>
 <19991102154543.A4469@optiplex.palga.uucp>
 <381F9EB4.63FA033F@python.net>
 <199911031734.JAA15489@hermes.aspectdv.com>
Message-ID: <4.2.0.58.19991115130555.00b8bba0@mail.aspectdv.com>

At 09:31 PM 11/12/99 -0800, Paul Eggert wrote:
>Jim Althoff <jima at aspectdv.com> writes:
>
> >Absolutely right.  Part of the original dream of the Smalltalk development
> >team was to build a programming language/environment that would be so
> >easy that anybody -- even kids -- could use it.
>
>Kids?  Kids??  You think _kids_ were the target!?
>No way.  The PARC developers were much more ambitious than that.
>Smalltalk's design goal was to be so easy
>that even _secretaries_ could use it.

I see.

Well, what can I say?  I was merely expressing my opinion based
on actually having been a member of the Smalltalk development
team at PARC.  (And, yes, we _did_ want secretaries to be able to
use the applications.)

Although I assume the "_secretaries_ vs _kids_" bit above was meant
to be humorous, for the historically curious, here is what Alan had to say:

 >>>>>>>>>>>>>>>>>>>>>>

Why is this collection of objects called Smalltalk?

The name actually appeared before the first Smalltalk design, which was 
created in 1971. I had mentioned to someone that the "prose" of then 
current programming languages was lower than a cocktail party conversation, 
and that great progress would have been made if we could even get to the 
level of making "smalltalk". It then occured to me that this word was also 
a great pun for a language aimed at the world of small children.

As for hype: Another motive for "smalltalk" was the practice at the time of 
naming operating, and other, systems (which hardly did anything) after 
mighty indo-european gods, such as Thor, Zeus, Odin, etc. I figured that if 
Smalltalk ever did something neat, then people would be pleasantly surprised.

As related in the Early History of Smalltalk (in HOPL II, Addison Wesley, 
1995), the actual start of today's Smalltalk was kind of accidental and was 
partly the result of a bet. -- Alan Kay

<<<<<<<<<<<<<<<<<<<<<

Jim






>--
>http://www.python.org/mailman/listinfo/python-list





From dkaznadzey at my-deja.com  Mon Nov  8 10:57:43 1999
From: dkaznadzey at my-deja.com (dkaznadzey at my-deja.com)
Date: Mon, 08 Nov 1999 15:57:43 GMT
Subject: Popup menus in Tkinter
Message-ID: <806rtk$4e6$1@nnrp1.deja.com>

Does anyone know weather there is a way to create
a popup menu in tk (using Tkinter)? It seems
like in order to be shown the menu must
grow from the valid menubar. If there is a way
to display standalone menu pane at given
coordinates, please let me know.

Denis Kaznadzey


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



From boud at rempt.xs4all.nl  Tue Nov  2 15:56:21 1999
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 2 Nov 1999 20:56:21 GMT
Subject: pygettext
Message-ID: <7vnj5l$ha$1@essle.valdyas.org>

I'm trying to write a resources page to go with my Python/KDE tutorial,
and I want to point people to pygettext.py, which I found somewhere
on my harddisk. However, there isn't a copyright notice in the file,
and I'd like to mention the author, too. Could he please contact me?
I've had to adapt the program slightly to work with the kde i18n
function, too.

-- 

Boudewijn Rempt  | http://www.xs4all.nl/~bsarempt



From vincent at sqr.nl  Wed Nov  3 09:15:11 1999
From: vincent at sqr.nl (Vincent Hillenbrink)
Date: Wed, 03 Nov 1999 15:15:11 +0100
Subject: MySQLmodule mainainer
Message-ID: <3820436F.136B@sqr.nl>

Does anybody know the e-mail addres of the current MySQLmodule
maintainer? (senekow at ibm.net bounces). I've made a modification
to the module which I'd like to see as a part of the next release.

Thanks,
Vincent



From grant.griffin at iowegian.com  Mon Nov 15 18:16:24 1999
From: grant.griffin at iowegian.com (Grant Griffin)
Date: Mon, 15 Nov 1999 23:16:24 +0000
Subject: OT: wide-open software (was: python)
References: <3820C0B8.62B2@iowegian.com> <3821953B.4BDC698C@mitre.org> <38221B4D.FA4@iowegian.com> <80kvt2$3ig$1@nntp6.atl.mindspring.net>
Message-ID: <38309448.33E9@iowegian.com>

Fred Marshall wrote:
> 
> Grant,
> 
> I may be missing something here with regard to the WOL.  How is it different
> from the BSD, NPL, MPL, X, Python or Apache licenses or just Public Domain?
> All of these allow modifications to be taken private.  Using what's already
> widely understood and published (and written about widely) seems a good
> idea.  Not that yet another is so awful either!  Creating new ones seems
> more the work of lawyers needing to earn their keep (in perception).
> 
> Fred Marshall
> 

Hi Fred,

The "Wide-Open License" (http://www.dspguru.com/wol.htm) is a little
different from Public Domain because with PD, you give up all rights to
your software: you can't place conditions on its use.  The WOL (and
maybe all of the licenses you mention) does impose the very minimal
condition that the author's name copyright notice remain.

But beyond that, you're absolutely right that the WOL has a very similar
intent to those on your list.  It probably even uses many of the same
words.  (I didn't actually invent the WOL text, and I don't know who
did: it's been kicking around for many years.)

The primary difference, then, is simply one of "marketing".  The WOL
isn't just a license: it's a slogan and a brand name.  Open-source
software authors already know what "copyleft" means, but until now there
has been no term for the converse.  Since it's a lot easier to think and
talk about a concept if there's a word for it, I coined the term
"wide-open" to mean "free of cost, open source, but not copylefted".

The wide-open concept carries "open source" to its logical conclusion:
if open is good, wide-open is better.  As you might guess, then, the
_raison d'etre_ of the WOL is to compete with the GPL (GNU Public
License) in the minds of software-license "consumers" (that is, software
authors.)  The WOL gives them a name-brand alternative to the GPL.

Now...as it turns out, Fred, your list actually illustrates a couple of
benefits of the WOL. :-)

First, each of the licenses you name are associated with a specific
open-software product.  Therefore, they carry whatever associations
consumers have of those products; for example, people who prefer
GNU/Linux to FreeBSD might be reluctant to use the "BSD" license.  Yet
the BSD license itself stands on its own; you could change a few words
and then use it for anything.  The solution, of course, is simply to
create a license which is named for the primary feature of the license:
that it's wide-open.

Your list also illustrates the need for a "brand name" for
wide-openness.  Prior to the WOL, open-software licenses consisted of
the "GPL" (embodying copyleft) and "other", including many similar
wide-open-style licenses, plus a few "oddball" licenses (e.g. Gnuplot
and Scilab).  (Editorial Aside: If the well-intended branding effort of
Opensource.org is wildly successful, it will result in the creation of a
large number of new open-source licenses, further confusing both authors
and users.)  

While many people truly subscribe to the copyleft ideology, I suspect
that some authors of free/open software use the GPL simply because it's
easier than choosing from among the alternatives.  Or maybe they think
that if you're writing open/free software, you're just "supposed to" use
the GPL.  So the fact that you were able to name 6 similar licenses
suggests that we need a _single_ wide-open license to focus on to reduce
consumer confusion.  At the risk of confusing consumers with a 7th, I
have created the WOL.

Unfortunately, to fully compete with "copyleft" and the GPL, the WOL has
to be part "ideology", so here goes.  On the psychological side, I hope
authors will feel good about giving their "software gift" without
limitation; this is the best spirit of giving.  On the practical side, I
hope authors will recognize that being open-minded about who can adopt
their software child tends to encourage the poor little thing's
adoption.  (And doesn't it deserve a nice, warm home?)

Python's creator, Guido van Rossum, has long provided leadership in
wide-openness by licensing Python under its own wide-open-style
license.  I don't know whether this is an expression of his personal
values or simply a very pragmatic recognition of the fact that wide-open
software has an inherent "competitive advantage" over copylefted
software (Guido's a sharp guy), but I'm glad to see that much of the
Python community seems to be following his lead by reusing Python's
licensing terms.
  
Next, I'd like to see the openness that's embodied in these licenses
spread far and wide.  And  I think that a little bit of marketing is the
next logical step to open that up.  Using the "wide-open" slogan and the
"WOL" brand name, I'm trying to get people to buy into the wide-open
concept, and to popularize the WOL as the recognized alternative to the
GPL.  Then, software authors can choose between the two licenses
according to whether their intent really is to guard the perpetual
"free" status of their software (by using the GPL) or to maximize the
"gift value" of their work and to "encourage the adoption of their
software child" (by using the WOL).

=g2
-- 
_____________________________________________________________________

Grant R. Griffin                                       g2 at dspguru.com
Publisher of dspGuru                           http://www.dspguru.com
Iowegian International Corporation	      http://www.iowegian.com



From aj10 at my-deja.com  Tue Nov 30 18:19:58 1999
From: aj10 at my-deja.com (aj10 at my-deja.com)
Date: Tue, 30 Nov 1999 23:19:58 GMT
Subject: XPython (X11)
References: <383E5350.1E07536A@xrce.xerox.com>
Message-ID: <821m2h$7ne$1@nnrp1.deja.com>

In article <383E5350.1E07536A at xrce.xerox.com>,
  Emmanuel Pietriga <emmanuel.pietriga at xrce.xerox.com> wrote:
> Is anyone using XPython around here?
>
> I need help, or at least links to examples (I haven't found any but
the
> ones within the package)
>
> Thanks.
> Emmanuel
>
>
Python over Tcl/Tk - which runs over XWindows - is well supported and
you will find lots of good examples, experience and support. Also this
way you would make your app more reusable. So why do you want XPython ?
Graphics performance is critical or something?

Just curious.

Thanks
aj



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



From kcazabon at home.com  Tue Nov 16 00:15:55 1999
From: kcazabon at home.com (Kevin Cazabon)
Date: Tue, 16 Nov 1999 05:15:55 GMT
Subject: win32api... another question:  GetEnvironmentVariable("PATH")
References: <1269438868-3874906@hypernet.com>
Message-ID: <fM5Y3.69771$Rx2.824379@news1.rdc1.bc.home.com>

Thanks... I'll have to go that route I guess.  I was hoping that if you
could access the SetEnvironmentVariable Win32 function, you could modify the
PATH system wide... it turns out not to be the case anyways, so the
os.environ route will work fine.

Thanks for the input...

Kevin.
Gordon McMillan <gmcm at hypernet.com> wrote in message
news:1269438868-3874906 at hypernet.com...
> Kevin Cazabon wrote:
>
> > Interesting... it doesn't work on Win98 for me (The exact statement
> > below...).
> >
> > Anyone have suggestions for Win95/98?
> >
> >
> > Kiss, Arpad <AKiss at GEOMETRIA.hu> wrote in message
> > news:01BF2F61.1D2B8DB0 at exchange1.geometria.hu...
> > > Hi,
> > >
> > > Kevin Cazabon <kcazabon at home.com> wrote in message
> > > news:<fcMX3.65012$Rx2.773564 at news1.rdc1.bc.home.com>...
> > > > I can't seem to get the win32api.GetEnvironmentVariable() function
to
> > > work
> > > > properly... am I not using it correctly by the following?
> > > >
> > > > path = win32api.GetEnvironmentVariable("PATH")
> > >
> > > It works for me on a Windows NT Workstation (sp5).
>
> First, why not use os.environ['PATH']?
>
> Second, while I'm on NT and 'PATH' and 'path' return the same
> answer (both the os and the win32api way), I had a bug report
> from someone on Win9x and the solution turned out to be:
>  try:
>    value = os.environ[key]
>  except KeyError:
>    value = os.environ[string.lower(key)]
>
>
>
> - Gordon
>
>





From phd at phd.russ.ru  Sat Nov  6 07:17:34 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Sat, 6 Nov 1999 12:17:34 +0000 (GMT)
Subject: Get parameter by name
Message-ID: <Pine.LNX.4.20.9911061213410.18974-100000@fep132.fep.ru>

Hello!

   I have a function with a long list of arguments:

def f(a1, a2, a3, b1, b2, b3, td, sd, si, ka, num=0, date=None):
   ...body_here...

and I have a list of strings "a1, td, ka". How can I get the value of
corresponding args? E.g., if I have

   x = "ka"

how can I get the value of parameter, the name of which is in x ("ka", in
this case)?

Oleg.
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From garry at sage.att.com  Tue Nov 30 10:05:22 1999
From: garry at sage.att.com (Garrett G. Hodgson)
Date: Tue, 30 Nov 1999 15:05:22 GMT
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <81g8po$reo$1@mach.vub.ac.be>
Message-ID: <3843E7B2.47ACC1F5@sage.att.com>

Thomas Hamelryck wrote:

> Well, here are some of the things I don't like very much:
> 
> Block delimitation by indenting.

this is perhaps my favorite thing about python, 
the lack of visul cruft.  when i look at code, i want to see
code, and not a bunch of noise that only serves to make the
parser writer's job easier.
(and yes, the last few languages i wrote used ";" and "{}".
but i plead ignorance, that was pre-python).

> This makes python look
> like a porsche with a caravan attached :-).

i don't get it.

> I very much miss the unary arithmatic operations that C has (e.g. a+=1).
i got used to that pretty quick.

one thing i dislike, even while understanding the reasoning,
is the explicit "self" in member references.  no big deal, though.

> But of course, it's still a very nice language :-)

agreed.

-- 
Garry Hodgson			"Hey, mister, can ya tell me,
garry at sage.att.com		where a man might find a bed?"
Software Innovation Services	He just grinned and shook my hand,
AT&T Labs			"No", was all he said.



From akuchlin at mems-exchange.org  Tue Nov 30 14:40:36 1999
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: 30 Nov 1999 14:40:36 -0500
Subject: Removing null items
References: <821761$s8u$1@nnrp1.deja.com>
Message-ID: <3demd7hiob.fsf@amarok.cnri.reston.va.us>

sragsdale at my-deja.com writes:
> So I've got a very simple desire: to remove all false (I.E. None, 0, '',
> []) items from a list.

The filter() built-in function, when not passed a specific function,
does exactly this:

>>> filter(None, [ 1,2,0, None, [], '12', '' ])
[1, 2, '12']

filter(F, seq) returns all the elements of the sequence 'seq', for
which the function F returns true: for example, the following code 
takes all the strings of even length:

>>> filter(lambda s: len(s) % 2 == 0, ['a', 'ab', '', 'abcd'])
['ab', '', 'abcd']

If F is None, then filter() simply returns those elements that Python
treats as true.

--amk



From geoff at gnaa.net  Sun Nov 14 23:30:53 1999
From: geoff at gnaa.net (geoff nordli)
Date: Sun, 14 Nov 1999 20:30:53 -0800
Subject: NT admin scripts with python
In-Reply-To: <jEmX3.3257$Xo.17395@news-server.bigpond.net.au>
Message-ID: <002d01bf2f22$34a16850$020a10ac@p350.gnaa.net>

I pulled some information from the ActiveState Perl site.  In the FAQ's they
tell you what kind of functions that you can perform.  There are many more,
but I only included the ones that I am interested in.

What modules come with the Perl for Win32 distribution?


Win32::Console: Perl module for Windows console interaction
Win32::EventLog: Read from and write to the Windows NT event log
Win32::File: Manage file attributes (read only, system, hidden...)
Win32::FileSecurity: manage ACLs in Perl
Win32::Internet: Perl module for Internet extensions
Win32::NetAdmin: Administer users and groups
Win32::NetResource: Manage resources (servers, file shares, printers)
Win32::Process: Start and stop Win32 processes
Win32::Registry: Read and manage the Win32 Registry
Win32::Service: Manage Windows NT services

It looks like Python can handle the NetAdmin module.  Are there other
modules that are available for Python.

thanks,

geoff
-------------------------------


Yep - the win32net module is all you need!  Included in the win32all
extensions...

Mark.

geoff nordli wrote in message
<000101bf2e22$7060a190$020a10ac at p350.gnaa.net>...
>Does python have a module that would allow a person to write
>scripts on NT Server to lets say:  create users, map drives,
>modify registry, create shares, and so on.
>
>thanks,
>
>Geoff Nordli
>





From skip at mojam.com  Thu Nov 18 12:53:17 1999
From: skip at mojam.com (Skip Montanaro)
Date: Thu, 18 Nov 1999 11:53:17 -0600 (CST)
Subject: UserDict, UserList... UserFile?
In-Reply-To: <19991118120859.A2974@optiplex.palga.uucp>
References: <19991118120859.A2974@optiplex.palga.uucp>
Message-ID: <14388.15629.683370.752256@dolphin.mojam.com>

    Gerrit> why isn't there a UserFile class, while there is a UserList and
    Gerrit> a UserDict?  And what about other builtin-types for which this
    Gerrit> is possible?

I suspect because nobody's ever wanted one badly enough thusfar.  I wrote a
simple UserNum module several weeks ago, but deleted it before posting it.

Skip Montanaro | http://www.mojam.com/
skip at mojam.com | http://www.musi-cal.com/
847-971-7098   | Python: Programming the way Guido indented...




From aruth at intercation.com  Tue Nov 30 15:15:48 1999
From: aruth at intercation.com (Adam Ruth)
Date: Tue, 30 Nov 1999 13:15:48 -0700
Subject: CGI Disconnect
Message-ID: <38443074.26D3F1C9@intercation.com>

I'm new to Python and Apache, my experience is using ASP with IIS.  Is
there a way to tell, using CGI, that the client is no longer connected?
ASP has a ClientConnected property that makes it easy to stop a long
query if the client has stopped waiting.

Thanks for your help,

Adam Ruth




From juri at iki.fi  Thu Nov 25 21:07:02 1999
From: juri at iki.fi (Juri P Pakaste)
Date: 26 Nov 1999 04:07:02 +0200
Subject: rfc822 and Received
Message-ID: <8fvyabm0zw9.fsf@sirppi.helsinki.fi>

The rfc822 module drops all but the last instance of headers that
appear multiple times in the message; a prime example of this is the
Received header (I'm not even sure if there are any others.) I don't
see any way to work around this particular feature, or how to fix it
without breaking compability. How do people handle them?

-- 
[ Juri Pakaste / juri at iki.fi / <URL:http://www.iki.fi/juri/> ] 



From T.Middletan at news.vex.net  Mon Nov 15 12:00:24 1999
From: T.Middletan at news.vex.net (T. Middletan)
Date: 15 Nov 1999 17:00:24 GMT
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us>
Message-ID: <80pe77$osc$1@hub.org>

[ Hmmm, i replied to this message already earlier via the python-list
eGroups web interface... but it has not shown up in usenet, and also
is not listed in the eGroups new message list. So something seems
to have messed up somewhere. Just as well, it was a somewhat
annoyingly windy message, and terribly self depricatory, so just
as well it didn't make it and i'll edit it a bit and try again.
Oh dear, starting out committing the same crimes again already!
Oh well! Incorrible! ]

>The issue of indexing modules is an old one, and I'm still astounded
>that we've never converged on a solution, though everyone always
>complains about this.

Yes, i'm astounded by it too. I've been skulking around c.l.py for
some time, and I've seen the occasional flare-ups, and proposals, and
tries, and abandoned projects...

One day when i was looking desperately for a resource that I remember
seeing posted in the newsgroup and not being able to find it, i
snapped... and maniacally (if not furtively) thought i'd see if i
could put something together. I figured it'd be useful for my own
archival purposes if nothing else! And the result is Parnassus, such
as it is.

I resisted announcing it for many reasons. I wasn't sure i'd be able
to pull it off is one of them. Another is that I didn't want to be
another of the voices crying in the wilderness over this, and not
get anything done. And the rest of the reasons I'll skip, lest this
message become too windy and self-deprecatory!

So that is the story of that.

>and the Locator-SIG was started in 1995.  Looking through the SIG

Hmm, i missed that SIG. Will have to have a look.

>The backbone of Trove (http://www.tuxedo.org/~esr/trove/) is partially

I'd never heard of this until Jules mentioned it either. Very
interesting. I shall say more shortly.

>The problem is getting acceptance from the community; people need to

Yep. Which is another reason my project has an initial advantage in
that it was designed "for my own use, if nothing else"... i'm not
waiting or relying on acceptance. I have been stocking and updating
the database myself (so if nothing else, at least *I* can find things).
Before today only one other person had ever submitted anything. (Today
there was a small flood which i wasn't quite prepared for! But i've
managed to cope with it, it would seem <G>).

>always, *always* release things on CPAN, and follow CPAN's naming scheme.

Originally i had a longer more vitupritive rant here, but lets skip
it this time! Oh what the hell. Everyone skip the next paragraph
please. Thanks.

<personal rant> I hate CPAN. I have always hated CPAN. I will always
hate CPAN! All right, okay, in truth back when i was a PERL hacker i
thought CPAN was the greatest, very convenient, it's true. But after
the great light of Python descended upon me (actually it was kind of
thrust upon me as a "strong suggestion", at a time when i had no
interest in Python whatsoever --- thank D'Arcy of PyGreSQL fame!) my
perspective changed. I realised PERL was amazingly painful to code,
and even more painful to maintain --- why didn't i realise this
before? And as the scales fell from my eyes i realised that CPAN was
incredibly ugly and tedious to wade through, now that i thought about
it. A sea of README files, and ugly filenames.

My apologies to CPAN maintainers. Opinions may differ.

>We just need to push the Python community very hard to enter their data.

The current volume of python releases isn't that huge. At the moment
I don't mind (and have not minded) passively watching and snagging
everything manually. Though there is of course a fairly decent (though
could be improved) submission form at Parnassus ready and waiting for
anyone, which would make things even easier---although not quite
automated---at this end.

>seeding the initial database, and that's important, because it makes the
>index immediately useful.

My feelings exactly. That's the goal. To be useful. Immediately is the
best way to be useful.

>Yet I hadn't heard about it until now; is the
>site still beta?  Did I miss a c.l.py.announce article about it?

Oops, i guess i forgot to announce it. <-;

It's been more or less running for 2 or 3 weeks, maybe a month. With
no real announcement. Seems fairly stable. Had only gotten about 40
hits in that time. But since yesterday around noon i've gotten about
200. (Uh, this is up to about 300 now, since my first writing of this
message. Whee).

There are a lot more ideas and things i would like to add/change about
the site. Nothing is set-in-stone, so to speak. But i think it's
fairly stable and useable as it is now.

>There are various issues with it -- I don't know how scalable the
>underlying code is,

Well, yes, there are issues. But i'm not sure which exact issues you
are referring too. So rather than slagging my own site (as i'm very
tempated to do, and did in my first post!), i'll wait for complaints.
(-: Or suggestions.

One of the issues is (i must moderate my language here) that I'm
not a "real" programmer. I hope no one holds this against me. (-:
I'm not in any way trained in computer sciences or database
management. I'm just a self-taught hobbiest. It makes for somewhat
messy code, and lots of trial-and-error approaches.

What i'm saying is that i don't know exactly what you mean when you
say "scalable".

I'll just mention a few details, perhaps you can glean something from.

The data is contained in a relational database (being served currently
by PostgreSQL. The design structure is (i think) extremely flexible.
The "tree" system can be easily modified, and can be infinitely
deep (though of course this is not necessary). All objects can be
nested within any number of other objects in the database.

In fact when i went and looked at the Trove proposal afte hearing
about it from Jules i was surprised to see that in many ways my
database is very similar to that proposal. I think I could move
in the direction of compliance with it over time without too many
migraine headaches.

The code itself is much sloppier, but reasonable "modular" and
flexible (much of the main display code is based on a template system
of my own contrivance, for example). It can be expanded and modified
without insurrmountable difficulty at least, in my estimation.

However, yes, as you say above, and i already agree, there are
issues... inteface issues, design issues, all sorts of issues.

>the graphic design could be simpler,

Heh! I will say only that my main goal in design was to make it as
fast and easy to browse and search, as possible. My secondary goal
was, in reaction to the sterile environment of CPAN, to create some
"atmosphere". I may have gone a little overboard with that. But i
don't appologise... i like it. But alas, it may not look exactly
"professional", i admit. But it's functional despite that!

Of course it doesn't look very fabulous in Lynx. But even there it is
still fully functional... just a bit hard on the eyeballs.

>vex.net's reliability and bandwidth are unknown,

Yes, this is a concern to me also. It's not my server. I'm just a poor
fellow sitting in a basement with his old p166. I just have a lowly
shell account on vex.net. I have bandwidth quotas. About 100 megs a
week. I have absolutely no clue how much traffic Parnassus might
generate if it caught on... this is another reason i have been
hesitant to announce it. Perhaps someone might give me an estimate for
this?

As for realiability of the server itself, bandwidth aside... i don't
have too many fears there. D'Arcy is always making improvements. But
then again, i really have no real conception as to the potential load.

Here's what we are running at Vex. http://www.vex.net/tech.py
(though i think some of this page needs updating... that'll be
somethign close to that, i imagine. 200MHz AMD K6 system with 128M
RAM. Running Apache with PyApache at the moment. No persistant
database connection.

Maybe D'Arcy will notice the vex.net logo i put at the bottom of all
the pages and give me a break on the bandwidth and beef up the system
a little (he has been looking into some other alternatives to PyApache
lately I believe). (-: [ <mini-plug> Vex.net is a nice little server
and does offer relatively inexpensive shell accounts with Python (and
now Zope) installed! ]

>and we can argue about the classification hierarchy *forever* --
>but it's still valuable.

Oh, please, please ... argue it with me! Heh. Seriously, if anyone
wants to suggest better categorisations i'm *very* interested. As
mentioned above, i'm not all that educated in many aspects of
programming --- i've done the best I could (starting off with copying
the categories on the python.org Contrib page)... but i'm sure it
could be much better. The database is very flexible and changing
things around is not a problem in this respect.

I have no ego when it comes to this stuff! No one need worry about
offending my categorisation sensibilities. I actually explicitly
placed the "suggest categories" link beside the details view for
each item to cover up for these sins of ignorance of mine. I'm hoping
if people who know more see things in the wrong places they will be
swift to take advantage of the convenient form and let me know.

The one thing which has kept me from worrying about the
classifications *TOO* much is that i figure most people will rather
"search" for what they want than to "browse" for it. So hopefully
people can find what they are after that way, even if the
classification is a bit dodgy. The search engine could be beefed up
quite a bit too. I have many ideas of how to make it more efficient
and accurate... maybe in time.

>and has the incalculable advantage of being implemented.

Yes, that's my feeling as well. Perhaps Parnassus's only advantage.
Although "incalculable" may be a bit overstated! I'm certainly open to
anyone else's ideas though. I'm just taking things as they come, and
seeing how it all goes.

>Is everything you've done in the index?  If not, add the missing items.

Yep, everything is pretty much up to date, as far as data. What you
see is what you get. (Though i still do have a few hundred of the
oldest c.l.py.ann messages i occasionally browse through... but these
are 6 months or more old...) I think i've managed to enter just about
every reference which has been mentioned in c.l.py in the last few
months... and c.l.py.ann for longer than that. Of course i ripped
everything from the python.org Contrib page long ago. (-:

>Can we prominently link to it from python.org, so people begin to use it?

You could do that. Though i find the prospect slightly frightening.
And we may run into bandwidth issues. I'll be keeping an eye on things
fairly closely. If they get too out of hand i may have to ... i don't
know what! Open the vault only on Sundays. <-; Or perhaps some other
solution can be found.

These then have been some of the whys and wherefores.

Your Crawling Parnassus Slave (who's true goal in life is merely to
one glorious day be able to sport such resplendent white whiskers as
those of the ever glowing R.D.---a few of whose novels i have indeed
myself read---as seen in his image off on A.M.K's site!),

T.





From mhammond at skippinet.com.au  Tue Nov 30 08:29:03 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Tue, 30 Nov 1999 13:29:03 GMT
Subject: Does Pythonwin support SDI. - A Newcomer to Pythonwin
References: <821fmv$4b5$1@news.vsnl.net.in>
Message-ID: <ziQ04.11108$VA6.52500@news-server.bigpond.net.au>

Umm - it _should_ :-)  Ive never done it, and dont know of anyone who has.
It does do dialog based apps which is sometimes sufficient.  Shouldnt be too
hard to get SDI going if it doesnt work...

Mark

ganesh wrote in message <821fmv$4b5$1 at news.vsnl.net.in>...
>I understand that pythonwin is an MDI application by itself. But I am
trying
>to develop a SDI application in Pythonwin and have no idea where to start
as
>there is no proper documentation or PyCSingleDocTemplate object in
>Pythonwin. Can anyone help me out.
>
>
>





From a.eyre at optichrome.com  Mon Nov  1 07:07:30 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Mon, 1 Nov 1999 12:07:30 -0000
Subject: PythonWin Debugger
In-Reply-To: <002401bf2457$f6500fe0$0501a8c0@bobcat>
Message-ID: <000201bf2461$abe80340$3acbd9c2@peridot.optichrome.com>

>> Does that include straight win32 (non-MFC) apps?
> Yes - it should work there.
Even apps with their own Win32 message loop?

> Eg, if you run "pythonwin\pywin\debugger\fail.py" as a normal
> script from Python.exe, the debugger should fire up correctly.
I still get the same error from my embedded app. (Not using TK)

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From a.eyre at optichrome.com  Wed Nov  3 09:39:23 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Wed, 3 Nov 1999 14:39:23 -0000
Subject: NT DLL modules and C++
In-Reply-To: <7vnovm$6do$1@sun500.nas.nasa.gov>
Message-ID: <002701bf2609$38622320$3acbd9c2@peridot.optichrome.com>

> How do I get Python to call the initialization under NT?

This should do the trick (for module spam.dll):

#include <Python.h>

extern "C" void __declspec(dllexport) initspam()
{
	// Initialisation code here...
}

Make sure you link with python15.lib

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From kevinmcd at linuxsolutions.NOSPAM.demon.co.uk  Sun Nov 14 12:25:09 1999
From: kevinmcd at linuxsolutions.NOSPAM.demon.co.uk (Kevin McDermott)
Date: Sun, 14 Nov 1999 17:25:09 GMT
Subject: Python and sending emails with it
References: <382E2DB2.A00B9109@nv2.cc.va.us>
Message-ID: <01bf2ec5$397272a0$0202a8c0@kevinmcd>

I have a small function to do just this on my site...

It provides for a file to be sent as a non-attachment, in my example it is
instructions
for the data that is attached...

It's used in an application that I developed for my employer, but the code
is free.

http://www.linuxsolutions.demon.co.uk/

Kevin
Jeremy Misavage <tekhir at yahoo.com> wrote in article
<382E2DB2.A00B9109 at nv2.cc.va.us>...
> I'm kinda new to python and I'm trying to write a small program that will
take
> in form data, save it in a db, and then send an email to a person.
> 
> The program processes a users request for a "equipment request."  It gets
> loggged in a db (already done) and sends an email to the person in change
of the
> equipment.  I need help on the email part.  Does some out there have some
code
> that will taken data and send it to a person.  It shouldn't need a
username and
> password, but if it does I'll live with that.
> 



From dgrisby at uk.research.att.com  Fri Nov  5 09:41:12 1999
From: dgrisby at uk.research.att.com (Duncan Grisby)
Date: 5 Nov 1999 14:41:12 -0000
Subject: C++ extension modules on NT
Message-ID: <7vuqa8$fvi$1@pineapple.uk.research.att.com>

I am working on a binding from omniORB, a free CORBA ORB, to Python,
and I have run into a problem with Windows NT. The omniORB library is
written in C++ and, unfortunately for me, involves some global objects
with constructors. On Linux and Solaris, and presumably other Unixes,
these constructors are called correctly when Python loads my extension
module.

On NT, however, the constructors are never called, and the global
objects contain garbage. The document "Extending and Embedding the
Python Interpreter" mentions this problem in section 1.11, and says
that the problem doesn't occur if the Python interpreter's main()
function is compiled with C++. Unfortunately, this appears to be
incorrect, and the constructors still aren't called.

Does anyone have any experience with this sort of thing?  Is there
some way I can link a DLL so that it runs constructors when it is
loaded?  I've looked around on python.org, but I can't find anything.


By the way, if you want to try a pre-release of omniORBpy (but not on
NT...), see

  http://www.uk.research.att.com/omniORB/archives/1999-11/0027.html

for details.

Cheers,

Duncan.

-- 
 -- Duncan Grisby  \  Research Engineer  --
  -- AT&T Laboratories Cambridge          --
   -- http://www.uk.research.att.com/~dpg1 --



From homunq at my-deja.com  Thu Nov 25 01:53:36 1999
From: homunq at my-deja.com (homunq at my-deja.com)
Date: Thu, 25 Nov 1999 06:53:36 GMT
Subject: function as attr of subclass != method (???)
Message-ID: <81imde$3uk$1@nnrp1.deja.com>

I'm confused by the following. Is it a bug, or is something subtle going
on that I should understand?

>>> class parent:
	pass


>>> class child(parent):
	pass


>>> achild = child()
>>> parent.method = lambda self:"yup"
>>> achild.method()
'yup'
>>> achild.method = lambda self:"nope"
>>> achild.method()
Traceback (innermost last):
  File "<pyshell#102>", line 1, in ?
    achild.method()
TypeError: not enough arguments; expected 1, got 0


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



From tim_one at email.msn.com  Wed Nov  3 00:41:17 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 3 Nov 1999 00:41:17 -0500
Subject: How can I make my assertions smarter?
In-Reply-To: <381D7087.BF8510E5@webone.com.au>
Message-ID: <000601bf25be$0cdeb880$2e2d153f@tim>

[Stuart Hungerford, desperately hoping that design by contract will
 be easier in Python if he calls it by some other name ...]

> I'm using some simple functions to implement crude precondition,
> postcondition and invariant checking in Python classes.
>
> (Tim: please be gentle ;-)

I am always gentle with you, Stuart!  Your heart is in the right place --
although it may as well be firmly stuck up your ass for all the good it's
doing you <wink>.

> Is there some neat way to wrap these tests up so I can print any
> code expression that fails, e.g:
>
>      Precondition failed foo.py line 42: spoon.shape == 'bent'
>
> which would appear at line 42 of foo.py as
>
>      pre(spoon.shape == 'bent')

As Michael Hudson amply demonstrated, you're fighting the language.  If you
can't live with

    assert spoon.shape == 'bent', "precondition failed"

then, if I were you, I'd write a little preprocessor to transform your
pre/post/invariant(...) thingies into the desired code.  That's fighting the
language too, of course, but in a way that's easier to extend along with
your ambitions.  DBC runs deep!  I don't think it can be tacked on
satisfactorily.

generously-leaving-eiffel-a-reason-to-exist<wink>-ly y'rs  - tim





From alan.gauld at gssec.bt.co.uk  Tue Nov  9 12:59:36 1999
From: alan.gauld at gssec.bt.co.uk (Alan Gauld)
Date: Tue, 09 Nov 1999 17:59:36 +0000
Subject: Newbie question: Input
References: <3826b693.3888197@212.225.36.99>
Message-ID: <38286108.D30CF0BC@gssec.bt.co.uk>

> Whats the syntax to prompt for user input?

Try my tutor at:

http:/members.xoom.com/alan_gauld/tutor/tutindex.htm

Look at the 'Talking to the user' section.

Alan G.
-- 
=================================================
This post represents the views of the author 
and does not necessarily accurately represent 
the views of BT.



From smigowsky at dspace.de  Thu Nov 11 02:38:41 1999
From: smigowsky at dspace.de (Stefan Migowsky)
Date: Thu, 11 Nov 1999 08:38:41 +0100
Subject: Win32api.... can it create a Windows shortcut (.lnk) file?
Message-ID: <137053459BA7D211B3930090272A62CA8EDD80@newsintern.dspace.de>

Hi 


Mark Hammond did a great job on wrapping the win32api. So 
creating a shortcut is just so easy :
################ CreateShortCut.py ####################

from win32com.shell import shell
import win32api
import pythoncom
import os
import sys

def CreateShortCut(Path, Target,Arguments = "", StartIn = "", Icon = ("",0),
Description = ""):
    # Get the shell interface.
    sh = pythoncom.CoCreateInstance(shell.CLSID_ShellLink, None, \
        pythoncom.CLSCTX_INPROC_SERVER, shell.IID_IShellLink)

    # Get an IPersist interface
    persist = sh.QueryInterface(pythoncom.IID_IPersistFile)

    # Set the data
    sh.SetPath(Target)
    sh.SetDescription(Description)
    sh.SetArguments(Arguments)
    sh.SetWorkingDirectory(StartIn)
    sh.SetIconLocation(Icon[0],Icon[1])

    # Save the link itself.
    persist.Save(Path, 1)

if __name__ == "__main__":
	TempDir = os.environ["TEMP"]
	WinRoot = os.environ["windir"]

	Path        =  WinRoot    + "\\Profiles\\All Users\\Desktop\\New
Link.lnk"
	Target      =  Pythonroot + "pythonw.exe "
	Arguments   =  TempDir + "\\test.py"
	StartIn     =  TempDir
	Icon        = (Pythonroot + "\\py.ico", 0)
	Description = "New Link"

	CreateShortCut(Path,Target,Arguments,StartIn,Icon,Description)


Stefan



Dipl.-Physiker Stefan Migowsky	
dSPACE GmbH				
Technologiepark 25			33100 Paderborn (Germany)
mailto:smigowsky at dspace.de 		http://www.dspace.de



-----Original Message-----
From: me [mailto:none at fake.net]
Sent: Thursday, November 11, 1999 7:54 AM
To: python-list at python.org
Subject: Win32api.... can it create a Windows shortcut (.lnk) file?


I'm just trying to find an easy to create shortcuts on Windows platforms
through Python... it doesn't seem to be a simple matter though. 
 I've gone
through all the Win32 docs, and found nothing, and have started writing
(with limited success) a C function to do it.

Is there an easy way?  q:]

Thanks,
Kevin Cazabon
kcazabon at home.com



-- 
http://www.python.org/mailman/listinfo/python-list



From bkenniso at worldnet.att.net  Mon Nov 29 18:30:31 1999
From: bkenniso at worldnet.att.net (Brian Kennison)
Date: 29 Nov 1999 23:30:31 GMT
Subject: Pthreads
Message-ID: <81v2an$n6o$1@bgtnsc01.worldnet.att.net>

I'm having trouble compiling Python 1.5.2. with the FSU pthread
library. I working on MachTen and get the following error:

gcc -g -O2 -I./../Include -I.. -DHAVE_CONFIG_H   -c thread.c -o thread.o
In file included from thread.c:150:
thread_pthread.h:113: parse error before `pthread_cond_t'
thread_pthread.h:113: warning: no semicolon at end of struct or union
thread_pthread.h:114: warning: data definition has no type or storage class
thread_pthread.h:115: warning: data definition has no type or storage class
thread_pthread.h: In function `PyThread_start_new_thread':       


thread_pthread.h looks like this:

   110 typedef struct {
    111         char  locked; /* 0=unlocked, 1=locked */
    112         /* a <cond, mutex> pair to handle an acquire of a locked lock */
    113         pthread_cond_t   lock_released;
    114         pthread_mutex_t  mut;
    115 }; pthread_lock;

There are more errors after this but I figure this is where it starts. 

Any help would be appreciated.

Brian Kennison




From hnowak at cuci.nl  Tue Nov 23 01:11:31 1999
From: hnowak at cuci.nl (Hans Nowak)
Date: Tue, 23 Nov 1999 07:11:31 +0100
Subject: Python books
Message-ID: <199911230613.HAA21179@dionysus.fw.cuci.nl>

A quick search at Proxis (online bookstore, http://owas.proxis.be) 
for Python yields, among others, the following books:

John Grayson: Python & Tkinter programming
Ivan van Laningham: Python in 24 hours
?: Python Annotated Archives

It claims that all these are published already. Does anyone know if 
this is really the case? Proxis has been known to put "published" 
books in their list long before they actually hit the shelves.

--Hans Nowak (zephyrfalcon at hvision.nl)
Homepage: http://fly.to/zephyrfalcon
You call me a masterless man. You are wrong. I am my own master.



From wheinemanNOwhSPAM at uconect.net.invalid  Mon Nov  8 13:01:41 1999
From: wheinemanNOwhSPAM at uconect.net.invalid (wheineman)
Date: Mon, 08 Nov 1999 10:01:41 -0800
Subject: pythonistas or pythoneers?
References: <07111999.2@sanctum.jae>
Message-ID: <0a0133f8.2269f155@usw-ex0102-010.remarq.com>

How about pythonaut? Like Guido and the Pythonauts.

Willy



* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!




From rparker at bserv.com  Sun Nov  7 16:12:48 1999
From: rparker at bserv.com (Robert Parker)
Date: Sun, 07 Nov 1999 21:12:48 GMT
Subject: Printing from Python to Windows Printers
References: <fIHU3.641$f66.886@198.235.216.4> <804hb0$4cb6@SGI3651ef0.iddeo.es>
Message-ID: <kXlV3.180$bq1.131@198.235.216.4>

Thank you for the leads I will pursue and share my findings.

Juan <jlle at redestb.es> wrote in message
news:804hb0$4cb6 at SGI3651ef0.iddeo.es...
> You can use the class wxPrinter and wxPrintout instaling previously
> wxPython.
>
> Un saludo.
> Juan Huertas
>
> Robert Parker <rparker at bserv.com> escribi? en el mensaje de noticias
> fIHU3.641$f66.886 at 198.235.216.4...
> > I am trying to find out how I can print to a windows printer from a
Python
> > application.  Has anyone used any of the Com modules to gain access to
the
> > printer api in windows?  Are there any modules that have been written
that
> > would help save time when we want to send font information only to the
> > printer driver not have to write in the escape sequences and print
> directly
> > to the parallel port?
> >
> >
> >
> >
>
>





From m.faassen at vet.uu.nl  Thu Nov 25 22:11:50 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 26 Nov 1999 03:11:50 GMT
Subject: Python Type-Inference based LINT.. (pylint.py)
References: <19991121180043.B3184@teapot.egroups.net> <81e2r7$j14$1@newshost.accu.uu.nl> <383BC3D2.98BCD1FF@compaq.com>
Message-ID: <81ktpm$33g$1@newshost.accu.uu.nl>

Greg Ewing <greg.ewing at compaq.com> wrote:
> Martijn Faassen wrote:
>> 
>> Note that this doesn't need to be a syntax extension; check out some of
>> my Swallow posting with dejanews for examples of how it could work in
>> current Python.

> I think I'd prefer it if it *were* a syntax extension.
> The suggestions I've seen so far for non-syntax-extending
> type declarations strike me as rather ugly.

Oh, agreed -- it's just that it's potentially easier to whip up something
using basic Python, first. Not that much easier on the scale of things, of
course.

Imagine, for instance, a list of integers. A dictionary with string keys and
values of some class. How do you write that down in a proposed syntax 
extension? It's not hard to whip up something with basic Python lists and
dictionaries.

> That's a minor point, though -- the main thing is to
> get some sort of type checker to work! The syntax can
> be thrashed out later.

Agreed completely.

>> I myself would take the Python subset approach
>> (i.e. Swallow) to make sure no runtime dynamics exist that can mess up
>> any type checking. Seems to me the easier approach

> Me, too. Also, I tend to think that explicit type
> declarations are a valuable way for the programmer to
> state the *intent* of the code in a way that both
> humans and compilers can process. In my experience,
> type errors are found sooner and reported in a more
> easily understood way when explicit type declarations
> are used.

> I've tried it both ways when using Haskell-like
> languages, and I find that the error messages I get
> when I don't explicitly declare types tend to be of
> the form

>    Couldn't unify <some-big-hairy-complicated-type-
>    expression> with <some-other-big-hairy-complicated-
>    type-expression-that-looks-vaguely-similar-although-
>    there-must-be-some-subtle-difference-if-only-I-can-
>    find-it>.

> After getting a few of those I quickly learned to
> provide type declarations for ALL my functions...

Sounds like we'd need some way to avoid this type of problem with
any 'optional-type-annotated Python'. Of course the Swallow idea is to
*fully* type annotate everything, and then use this to build fast
extension libraries written in (a subset of) Python. This is just to make
things easier, though -- types would be useful in plain Python as well, 
if they were optional. I think. It's hard to say, actually, because if it
*has* to generate error messages like the kind you mentioned, optional
type annotations may not even be worth it..

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?



From henk_p at my-deja.com  Thu Nov 18 03:38:28 1999
From: henk_p at my-deja.com (henk_p at my-deja.com)
Date: Thu, 18 Nov 1999 08:38:28 GMT
Subject: Redirection
References: <80ui4k$m7u$1@news.fsu.edu>
Message-ID: <810du4$b5g$1@nnrp1.deja.com>

Instead of the normal cgi-header:

print 'Content-type: text/html'
print

use:

print 'Location: http://this.is.the/redirect/url'
print

That should do it.



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



From gmcm at hypernet.com  Mon Nov 29 12:12:12 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Mon, 29 Nov 1999 12:12:12 -0500
Subject: performance of tuple-less Python?
In-Reply-To: <HSx04.8911$cz3.321455@nnrp1.uunet.ca>
Message-ID: <1268226961-30285235@hypernet.com>

Greg Wilson asks:

> Has anyone ever built a version of Python that didn't include
> tuples, 

Not with the result of a working Python...

> and looked at the performance of code that consistently
> used lists instead?  

The relative performance would be related to the performance 
of malloc on your system. You could get some idea by timing 
creating and manipulating sequences of various sizes in 
regular Python.

> I'm curious as to whether the use of tuples
> (which I believe are conceptually redundant) 

They're only redundant if the concept of mutability is 
meaningless. A reference-based language without a concept 
of mutability would be, um, an adventure.

what-if-six-turned-out-to-be-nine-ly y'rs



- Gordon



From aa8vb at yahoo.com  Tue Nov 16 07:19:17 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Tue, 16 Nov 1999 07:19:17 -0500
Subject: Directory chooser in Tk
In-Reply-To: <38304E78.F92C8596@callware.com>
References: <38304E78.F92C8596@callware.com>
Message-ID: <19991116071917.A711753@vislab.epa.gov>

Ivan Van Laningham:
 |I need a dialog similar to the askopenfilename() dialog, but for
 |directories.  Is there such a beast?

There is a tk_chooseDirectory function in Tcl 8.2 (and in but undoc in
8.1), but reportedly this is Win32-only.

Christopher Nelson posted a proposed portable version of this
(chooseDir.tcl) to c.l.tcl last week.  There's also Daniel Roche's
tkgetdir.tcl.

A link to these posts along with a few others from DejaNews:

   http://www.deja.com/=dnc/[ST_rn=ps]/getdoc.xp?AN=547570099
   http://www.deja.com/=dnc/[ST_rn=ps]/getdoc.xp?AN=548332092
   http://www.deja.com/=dnc/[ST_rn=ps]/getdoc.xp?AN=532357262

Shouldn't be hard to port one to Python/Tkinter.  Alternatively, you could
modify the FileDialog.py module in the standard python dist to choose
directories.

-- 
Randall Hopper
aa8vb at yahoo.com



From news at dorb.com  Mon Nov  8 08:55:07 1999
From: news at dorb.com (Darrell)
Date: Mon, 8 Nov 1999 08:55:07 -0500
Subject: Newbie question: Input
References: <3826b693.3888197@212.225.36.99>
Message-ID: <rBAV3.9530$Xd2.521203@typhoon.nyroc.rr.com>

>From Python Library Reference:

raw_input ([prompt])
If the prompt argument is present, it is written to standard output without
a trailing newline. The function then reads a line from input, converts it
to a string (stripping a trailing newline), and returns that. When EOF is
read, EOFError is raised. Example:

>>> s = raw_input('--> ')
--> Monty Python's Flying Circus
>>> s
"Monty Python's Flying Circus"
>>>

If the readline module was loaded, then raw_input() will use it to provide
elaborate line editing and history features.

--
--Darrell
Endgamer <endgamer71 at hotmail.com> wrote in message
news:3826b693.3888197 at 212.225.36.99...
> Okay, I know it's a damn stupid question, but I've been through the
> books, the texts, the tutorial, the language reference and no luck...
>
> Whats the syntax to prompt for user input?
>
>
> Endgamer
> Spider Warrior
>
> (Endgamer71 at spamtrap.hotmail.com)
> #####################################
> Not a quote, just an on-the-spot observation:
> You know you're a geek when your PC costs exactly
> 15 times as much as your stereo. And gives better
> sound quality.<sigh>
>                               Endgamer
> #####################################
>
> --
> http://www.python.org/mailman/listinfo/python-list





From BjoernGam at knuut.de  Thu Nov 25 12:24:42 1999
From: BjoernGam at knuut.de (Bjoern Gam)
Date: Thu, 25 Nov 1999 18:24:42 +0100
Subject: What are the right rights for python scripts under Linux?
Message-ID: <383D70DA.5E4858BC@knuut.de>

Hello,

I have a simple question: "What are the >right< rights under Linux for
python scripts?" A WWW page send some informations to this python script
and the script send the request back to the browser. Which kind of
rights do you propose?

Thanks 
Bjoern



-- 
email: bjoerngam at knuut.de



From thomas at bibsyst.no  Fri Nov 12 04:01:22 1999
From: thomas at bibsyst.no (thomas at bibsyst.no)
Date: Fri, 12 Nov 1999 10:01:22 +0100
Subject: Object Persistence and Pickle&zlib
Message-ID: <382BD762.9E2502AA@bibsyst.no>

Hi,

I got a object declared like this ( this is an example, the actual code
is much bigger ) :

class some_object:
    title = ''
    id = 0
    items1 = {}
    items2 = {}
    items3 = {}
    items4 = {}

    def some_method1(self):
        ...

    def some_method2(self):
        ...

    def some_method3(self):
        ...

    def some_method4(self):
        ...

which I create like this :

x = some_object()
x.id = 2
x.title = 'test'

and use like

x.some_method1() # which fills up the dictionaries in the object with
various data
x.some_method2() # the same
x.some_method3() # lists out the data in an object for instance

and I store the object like

import shelve
db = shelve.open('objects')
db[str(x.id)] = x
db.close()

but when I get the object later, some methods and data seem to be
missing.
For example, the first three dictionaries in the object might have the
data I stored in them, but the last one is empty. Some of the methods
are available, but some are missing. Eh ... what is going on?

Do I have to initialize vars in any way to make them stick? The state of
the object is the same each time I fetch it from a datafile.

I also want to use compression on these objects, since the seem to be
reduced 1/10 in size. I do something like :

data = pickle.dump(x)
data = zlib.compress(data)

then store it like so :

db[str(x.id)] = data

When I try to fetch the object I get an -5 error from zlib :

data = zlib.decompress(db[id])

What does that error mean?

Thanks for helping a python-newbie out.

Thomas Weholt





From fdrake at acm.org  Thu Nov 11 17:50:38 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Thu, 11 Nov 1999 17:50:38 -0500 (EST)
Subject: WEB DATABASE PROGRAMMER POSITION AVAILABLE
In-Reply-To: <80fge0$rkv$1@hub.org>
References: <MuFW3.102$yd1.9175@typhoon-la.pbi.net>
	<80fdts$fml$1@nntp1.atl.mindspring.net>
	<80fge0$rkv$1@hub.org>
Message-ID: <14379.18494.226706.37109@weyr.cnri.reston.va.us>

D'Arcy J.M. Cain writes:
 > I thought we might all get together and send back a community resume with
 > all our experience and training.  :-)

  What?  And get us all squeezed into one little cubicle?  Let's break 
up into at least three or four competing clans!  ;-)


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From gmcm at hypernet.com  Wed Nov 17 17:50:11 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 17 Nov 1999 17:50:11 -0500
Subject: Beta 3e of the Win32 Installer
Message-ID: <1269243486-15626894@hypernet.com>

Win32 Installer users:

Beta 3e is here (the *last* beta before release 3, got that, 
Burt? <wink>).

The new beta features:
 - run from CDROM (if R/O, uses temp directory)
 - put your own icon in the .exe (thanks to Thomas Heller!)
 - bug fixes
 - more docs
 - examples of how to put Python on a floppy.

In fact, one of the configs will put Python, Tcl/Tk and IDLE on 
a floppy. Use it interactively, or pass it a script! Just like real 
Python.

Check out http://starship.python.net/crew/gmcm/install.html

A more formal announcement will follow when it goes release.

- Gordon



From gek8365 at garnet.acns.fsu.edu  Wed Nov 17 10:28:59 1999
From: gek8365 at garnet.acns.fsu.edu (Glenn Kidd)
Date: Wed, 17 Nov 1999 10:28:59 -0500
Subject: Subject line in SMTP
Message-ID: <80uhf9$m0d$1@news.fsu.edu>

I am writing a form mailer in python and I have completed all of it but I
have a question.  How can I get the subject line to appear on the message?
Any help would be appreciated.


Glenn Kidd

PS If you need to see the code let me know.





From Matija.Exel at lag.ensieg.inpg.fr  Mon Nov  8 04:18:41 1999
From: Matija.Exel at lag.ensieg.inpg.fr (Matija Exel)
Date: Mon, 08 Nov 1999 10:18:41 +0100
Subject: build failure for SunOS 5.5
Message-ID: <38269571.14471A7D@lag.ensieg.inpg.fr>

hello,

is there a way to build Python for SunOS 5.5? Here are my attempts
on a SPARCstation20 platform:

1) context information:
> cat /etc/motd
Sun Microsystems Inc.   SunOS 5.5       Generic November 1995
> which cc
/usr/ucb/cc

 2) configure and make:
Python-1.5.2>./configure
Python-1.5.2>make
"./posixmodule.c", line 629: undefined symbol: chdir
"./posixmodule.c", line 688: undefined symbol: chown
"./posixmodule.c", line 727: undefined symbol: link
"./posixmodule.c", line 1059: undefined symbol: rmdir
"./posixmodule.c", line 1130: undefined symbol: unlink
"./posixmodule.c", line 2844: warning: improper pointer/integer
combination: op "="
*** Error code 2
make: Fatal error: Command failed for target `posixmodule.o'
Current working directory
/home/etel/exel/HomePython/Python-1.5.2/Modules
*** Error code 1
make: Fatal error: Command failed for target `Modules'

When looking into congif.log, I find the following error messages:
"configure", line 1381: invalid source character: '#'
"configure", line 1381: warning: improper pointer/integer combination:
op "="
"configure", line 1381: non-constant initializer: op "NAME"
"configure", line 1381: invalid source character: '#'
"configure", line 1381: warning: improper pointer/integer combination:
op "="
"configure", line 1381: non-constant initializer: op "NAME"
"configure", line 1381: syntax error before or at: "() -- "
"configure", line 1381: warning: syntax error:  empty declaration
"configure", line 3887: warning: volatile is a keyword in ANSI C
"configure", line 3887: undefined symbol: volatile
"configure", line 3887: syntax error before or at: int
"configure", line 3887: undefined symbol: x
"configure", line 3913: warning: signed is a keyword in ANSI C
"configure", line 3913: undefined symbol: signed
"configure", line 3913: syntax error before or at: char

 any help appreciated!

--
-------------------------------------------------------------------
 M. Matija Exel
 E.N.S.I.E.G., Service  Informatique / L.A.G.
 BP.  46 Cedex,  38402 St.Martin d'Heres,  FRANCE
 Tel : (+33) 4 76 82 71 12          Fax:(+33) 4 76 82 63 88
 e-mail : Matija.Exel at lag.ensieg.inpg.fr
 web:  http://www-hadoc.ensieg.inpg.fr/~exel/





From kuncej at mail.conservation.state.mo.us  Fri Nov 12 15:02:24 1999
From: kuncej at mail.conservation.state.mo.us (Jeffrey Kunce)
Date: Fri, 12 Nov 1999 14:02:24 -0600
Subject: Shared NT network installation of python
Message-ID: <s82c1e30.070@mail.conservation.state.mo.us>

There was a bug in the initial upload that caused problems with the 
system path environment variable. The new, fixed version is available now.
If anyone was stung by the bug, the remedy is very simple - email me for details.

  --Jeff

>>> "Jeffrey Kunce" <kuncej at mail.conservation.state.mo.us> 11/10/99 05:50PM >>>
I've worked out a procedure to run python (including tkinter) on NT 
from a shared network installation.

It is working for me with only some minor glitches. I'm making it available 
with the ***CAUTION*** that it is an experimental work in progress. 
If it trashes your registry, don't blame me!

See:
    pynetinst
at:
    http://starship.python.net/crew/jjkunce/ 

Comments and suggestions are welcome.

  --Jeff







From bernhard at alpha1.csd.uwm.edu  Thu Nov 18 13:07:59 1999
From: bernhard at alpha1.csd.uwm.edu (Bernhard Reiter)
Date: 18 Nov 1999 18:07:59 GMT
Subject: How to put wxGrid into wxPanel ?
References: <38340b4e.71103058@news.dnet.net.id>
Message-ID: <slrn838g3t.64d.bernhard@alpha1.csd.uwm.edu>

On Thu, 18 Nov 1999 14:32:51 GMT, Ahmad Munif <muniff at yahoo.com> wrote:
> i have a problem to put wxGrid into wxPanel. the following is my code

Which version of wxPython are you using?
(I think I have 2.1b2 here, which is a bit old.)


And for me it works after giving the grid a default size:

from wxPython.wx import *
class MyFrame(wxFrame):
        def __init__(self,parent,id,title):
                wxFrame.__init__(self,parent,id,title,wxPoint(-1,-1),wxSize(300,200))
                panel = wxPanel(self,-1)
                grid  = wxGrid(panel,-1,wxPoint(10,10),wxSize(250,180))
                grid.CreateGrid(5,5)

class MyApp(wxApp):
        def OnInit(self):
                frame = MyFrame(NULL,-1,"wxPanel & wxGrid")
                frame.Show(true)
                self.SetTopWindow(frame)
                return true

app = MyApp(0)
app.MainLoop()



>from wxPython.wx import *
>class MyFrame(wxFrame):
>	def __init__(self,parent,id,title):
>		wxFrame.__init__(self,parent,id,title,size=(300,200))
>		panel = wxPanel(self,-1)
>		grid  = wxGrid(panel,-1)
>		grid.CreateGrid(5,5)
>
>class MyApp(wxApp):
>	def OnInit(self):
>		frame = MyFrame(NULL,-1,"wxPanel & wxGrid")
>		frame.Show(true)
>		self.SetTopWindow(frame)
>		return true
>
>app = MyApp(0)
>app.MainLoop()
>
>my code can't put wxGrid into wxPanel properly.
>but i can put others, such as button,textctrl,listbox, 
>into wxPanel properly (i can manage it)

Your parent seems to be fine.
You should have givin us the error message you get.
	Bernhard

-- 
Research Assistant, Geog Dept UM-Milwaukee, USA.  (www.uwm.edu/~bernhard)
Free Software Projects and Consulting 		         (intevation.net)  
Association for a Free Informational Infrastructure            (ffii.org)



From chrishmann at aol.com  Fri Nov  5 11:32:58 1999
From: chrishmann at aol.com (ChrisHMann)
Date: 05 Nov 1999 16:32:58 GMT
Subject: Python Conference Hotel Change
References: <14370.463.892526.967619@weyr.cnri.reston.va.us>
Message-ID: <19991105113258.01646.00002712@ng-ch1.aol.com>

Fred Drake writes:
>>Andy Robinson writes:
> > My memory may be fuzzy, but isn't it an even shorter stroll from there
> > to Bardo, one of the world's biggest microbreweries?
>
>  A big microbrew?  Strange... hope it's good!  ;-)

Sorry to disappoint, but Bardo's closed Oct. 31st.  I think they've moved out
to the suburbs of Fairfax or something like that.

-chris





From arad at fc.hp.com  Fri Nov 12 00:46:56 1999
From: arad at fc.hp.com (Arad Rostampour)
Date: Thu, 11 Nov 1999 22:46:56 -0700
Subject: print "foo", without a space
References: <382AD3BB.69899CBD@helsinki.xx>
Message-ID: <80ga4s$bks$1@web1.cup.hp.com>

Nothing with print comes to mind off hand, but you could just use
sys.stdout.write() instead (it wont insert a space).

Enjoy,
Arad
--
Arad Rostampour- arad at fc.hp.com
NT Platform and I/O Team
Software and System Development Lab
Hewlett-Packard

Mikael Johansson wrote in message <382AD3BB.69899CBD at helsinki.xx>...
>
>Hello All!
>
>Is there another way of getting print to not make a newline than
>using:
>    print "foo",  ?
>This because the comma always produces a space which necessary
>isn't that nice, for example:
>
>for i in range (1,30):
>    print i,
>    if i%10==1 and i<>11: print "st",
>    elif i%10==2 and i<>12: print "nd",
>    elif i%10==3 and i<>13: print "rd",
>    else: print "th",
>    print "fish"
>
>Now you would like the "endings" to come right after the number.
>There is naturally always some way to get around this, in this
>case for example:
>
>end=["st","nd","rd","th"]
>for i in range(1,30):
>    print `i`+end[max(min(abs(i)%10,4)-1-
>    4*(not(int('0'+`abs(i)`[-2:-1])-1)),-1)],'fish'
>
>OK, this got a bit out of hand, the above looks terrible, does
>anyone see how it could be simplified? At least it's more general
>than the above as it can handle the whole range of integers,
>including negative ones :-) However it shows that it would be
>less obfuscating to have a print statement that could continue
>printing on the same line without a space.
>
>If there isn't an easy way for this, I suggest the following:
>    print "foo".
>So that a dot at the end of the line would mean "no space either".
>
>Appreciating any comments.
>
>Have a nice day,
>    Mikael J.
>    for e-mail note that the country code for Finland is not xx





From philh at vision25.demon.co.uk  Mon Nov 29 16:03:46 1999
From: philh at vision25.demon.co.uk (Phil Hunt)
Date: Mon, 29 Nov 99 21:03:46 GMT
Subject: Htmlizing text
Message-ID: <943909426snz@vision25.demon.co.uk>

Is there a function in the standard Python library to HTML-ize text,
i.e. to replace 'a > b & c < d' with 'a &gt; b &amp; c &lt; d'?

(I realize this can be done with regular expressions, but is there
an easy way?)

-- 
Phil Hunt - - - phil at comuno.com




From godzilla at netmeg.net  Fri Nov  5 10:42:21 1999
From: godzilla at netmeg.net (Les Schaffer)
Date: Fri, 05 Nov 1999 15:42:21 GMT
Subject: ~ 1 = -2 ???
References: <000301bf275d$6468e120$1d2d153f@tim>
Message-ID: <m21za59beb.fsf@netmeg.net>

tim said bit-erly:

> Les, if x is always in range(16)

you bet!

class GroupElements:
    elements = range(16)
...


> simply subtract it from 15.  That is,

> def not4bit(x):
>     assert 0 <= x < 16
>     return 15 - x

> It's twice as fast <wink>.

for this particular app, its a little pseudo-group theory applied to
4-bit bit patterns. so  speed don;t matter. but i have some pixel op
stuff coming up, so i am glad i made some dumb mistakes NOW rather
than later!

why i would have even thought python's int's would be unsigned int's
is beyond me....

;-)

les schaffer



From prestonlanders at my-deja.com  Wed Nov 24 19:32:06 1999
From: prestonlanders at my-deja.com (Preston Landers)
Date: Thu, 25 Nov 1999 00:32:06 GMT
Subject: mysql and DATE format
References: <81hhm6$ba2$1@nnrp1.deja.com>
Message-ID: <81i027$lcp$1@nnrp1.deja.com>

Sounds like you need to either prefix your % symbols in the command with
a backslash, and/or use raw strings.

Try this:

"... DATE_FORMAT(date, \"\%W, \%d \%b \%Y\"), ..."

hope that helps. If that doesn't help, try prefixing the entire string
with the letter r and read the revelant sections of the documentation
(search for 'raw string' on the Python website.)


---Preston

In article <81hhm6$ba2$1 at nnrp1.deja.com>,
  sp00fD <sp00fD at yahoo.com> wrote:
> I've currently got some perl code that executes a sql statment which
> looks like:
>
> SELECT id, subject, author, text, DATE_FORMAT(date, \"%W, %d %b %Y\"),
> DATE_FORMAT(date, \"%h:%i %p\") FROM Table WHERE thread=18 ORDER by
> date desc
>
> When trying to use that in python like this:
> """SELECT .. DATE... FROM %s WHERE thread=%d""" % (table, id)
>
> it tries to expand the DATE %W.. variables.
> I've also tried it as
>
> sql = 'SELECT id, subject, author, text, '
> sql = sql + 'DATE_FORMAT(date, \"%W, %d %b %Y\"), '
> sql = sql + 'DATE_FORMAT(date, \"%h:%i %p\") FROM' + table
> sql = sql + 'WHERE thread=' + id + 'ORD...'
>
> which gives me a TypeError: illegal argument type for built-in
> operation.  Why?  How can I do this properly?
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
>

--
|| Preston Landers <prestonlanders at my-deja.com> ||


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



From fredrik at pythonware.com  Tue Nov 16 03:15:50 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 16 Nov 1999 09:15:50 +0100
Subject: Daemonic threads
References: <3830C73F.C8A7CA62@DONT.LIKE.SPAM>
Message-ID: <005c01bf300a$d06920a0$f29b12c2@secret.pythonware.com>

David S. <I at DONT.LIKE.SPAM> wrote:
> I'm writing a CGI script that will launch a long-running process in the
> background, and return some HTML page to the browser.  Now, I know how
> to do that with 'fork()', but it seems to me that I should be able to do
> the same thing with a daemonic thread from the 'threading' module.

daemon threads die when the python interpreter is shut
down, just like daemon processes die when the operating
system is shut down.  there's no way for a thread to out-
live the hosting process.

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
(news: just posted an errata for the 'first printing') -->




From a.eyre at optichrome.com  Mon Nov 15 10:10:41 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Mon, 15 Nov 1999 15:10:41 -0000
Subject: [Python] header (fwd)
In-Reply-To: <XFMail.991115154911.mikael@isy.liu.se>
Message-ID: <002b01bf2f7b$94dc4650$3acbd9c2@peridot.optichrome.com>

> Are you sure you don't mean "Sender:" instead of "From:"? I, at least,
> get the actual author at "From:", and "python-list-admin at python.org" at
> "Sender:".

Probably. I don't get that info using MS Outlook...

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From mal at lemburg.com  Fri Nov  5 06:24:03 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Fri, 05 Nov 1999 12:24:03 +0100
Subject: __str__ vs. __repr__
References: <000001bf275c$02370d20$1d2d153f@tim>
Message-ID: <3822BE53.A6DC6116@lemburg.com>

Tim Peters wrote:
> 
> David Ascher suggested that str(long) drop the trailing "L", and that's a
> good example.  The "L" is appropriate for repr(), for so long as Python
> maintains such a sharp distinction between ints and longs, but is really of
> no help to most users most of the time.

If you should really plan to do this, please add a new PyLong_AsString()
API, so that extensions can query the long int value using the
string format (probably the most portable way of passing long int
values from one implementation to another). A PyLong_FromString() API
would be a nice complement to PyLong_AsString()... it could use the
code from the builtin long() as engine.

I'm currently using PyObject_Str() on Python long integers to
have converted to strings and chop off the trailing 'L' "by hand".
Not really very elegant, but it works :-)

-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                    56 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/





From cjw at connection.com  Sun Nov 28 10:44:19 1999
From: cjw at connection.com (Colin J. Williams)
Date: Sun, 28 Nov 1999 10:44:19 -0500
Subject: A wxPython glitch?
Message-ID: <38414DD3.B2E8C625@connection.com>

The following method call gives an argument count error shown below.

  self.SetScrollbar(orientation= wx.wxHSCROLL,
           range= 700, thumbsize= 200,
           pagesize= 100, position= 0,
           refresh= wx.TRUE)

PythonWin 1.5.2 (#0, Apr 13 1999, 10:51:12) [MSC 32 bit (Intel)] on
win32
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
Portions Copyright 1994-1999 Mark Hammond (MHammond at skippinet.com.au)
>>> Unhandled exception while debugging...
Traceback (innermost last):
  File "G:\Program Files\Python\wxPython\windows.py", line 370, in
SetScrollbar
    val = apply(windowsc.wxWindow_SetScrollbar,(self,) + _args, _kwargs)

TypeError: wxWindow_SetScrollbar requires at least 5 arguments; 4 given
[Dbg]>>> _kwargs
{'range': 700, 'orientation': 1073741824, 'pagesize': 100, 'refresh': 1,
'thumbsize': 200, 'position': 0}
[Dbg]>>>

Any advice would be appreciated.

Colin W.




From Gaetan_Corneau at baan.com  Tue Nov 30 14:42:59 1999
From: Gaetan_Corneau at baan.com (Gaetan Corneau)
Date: Tue, 30 Nov 1999 14:42:59 -0500
Subject: WinNT and python as text filter...
Message-ID: <816010E2456BD111A48700805FBBE2EEFDF87A@ex-quebec-u1.baan.com>

Gordon,

> Are you sure it's the same problem?

Thinking of it, maybe not. 

 
> As far as I know
>  >xxx.yyy >foo.txt
> does not work if xxx.yyy is executed through a file 
> association.

On my machines, this *usually* works, because I use MKS Toolkit's Korn
Shell.
I tried with an NT "DOS" shell, and it doen't work. I wonder what's the
difference (no, I don't use #! in my script).


> foo.txt will be created and empty. The executable 
> associated to .yyy will have stdout and stderr connected to a 
> bit bucket. 
> Python doesn't have anything to do with it.

I know that, of course :)
______________________________________________________
   Gaetan Corneau
   Software Developer (Quality Assurance Team)
   BaaN  Supply Chain Solutions  
   http://www.baan.com
   E-mail: Gaetan_Corneau at baan.com                  
   Tel: (418) 266-8252          
______________________________________________________
"Profanity is the one language all programmers know best"





From gmcm at hypernet.com  Tue Nov  9 10:46:47 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 9 Nov 1999 10:46:47 -0500
Subject: Newbee WinNT ODBC Question
In-Reply-To: <809ckk$vbf$1@nnrp1.deja.com>
Message-ID: <1269960091-19324936@hypernet.com>

wayne_strauss writes:

> I've installed Python and Win32all.exe. When I run the following
> (only the first three lines shown):
> 
> import dbi, odbc     # ODBC modules
> import time          # standard time module
> 
> dbc = odbc.odbc('db/id/pw')
> 
> I get:
> 
> >>> Traceback (innermost last):
>   File "d:\Program
> Files\Python\Pythonwin\pywin\framework\scriptutils.py", line 237,
> in RunScript
>     exec codeObject in __main__.__dict__
>   File "D:\dev\Python\ODBC\odbc.py", line 1, in ?
>     import dbi, odbc     # ODBC modules
>   File "odbc.py", line 4, in ?
>     dbc = odbc.odbc('db/id/pw')
> TypeError: call of non-function (type module)
> 
> If the import worked then why does the actual function call fail?

You didn't by any chance name this script "odbc.py" did you?

import-Narcissus-ly y'rs

- Gordon



From parkw at better.net  Thu Nov 11 08:05:06 1999
From: parkw at better.net (William Park)
Date: Thu, 11 Nov 1999 08:05:06 -0500
Subject: .html -> .ps ?
Message-ID: <19991111080506.A1007@better.net>

Hi,

I need to convert HTML file to Postscript, so that I can send fax it.

1.  Does anyone have experience with 'html2ps', 'html2tex', and
    'html2latex' programs?  Or, is there any other program that will do
    the conversion in Python or otherwise?  

2.  Is there a command-line options to tell Netscape to print to a file?
    This would be ideal, since one can already do this manually with
    mouse.

	Yours truly,
	William Park



From juri at iki.fi  Thu Nov 25 21:15:49 1999
From: juri at iki.fi (Juri P Pakaste)
Date: 26 Nov 1999 04:15:49 +0200
Subject: rfc822 and Received
References: <8fvyabm0zw9.fsf@sirppi.helsinki.fi>
Message-ID: <8fvwvr60zhm.fsf@sirppi.helsinki.fi>

Juri P Pakaste <juri at iki.fi> writes:

> The rfc822 module drops all but the last instance of headers that
> appear multiple times in the message; a prime example of this is the
> Received header (I'm not even sure if there are any others.) I don't
> see any way to work around this particular feature, or how to fix it
> without breaking compability. How do people handle them?

Oops. Momentary blindness. Forget it. :-)

-- 
[ Juri Pakaste / juri at iki.fi / <URL:http://www.iki.fi/juri/> ] 



From mikael at isy.liu.se  Mon Nov  8 13:40:05 1999
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Mon, 08 Nov 1999 20:40:05 +0200 (MET)
Subject: Newbie question: Input
In-Reply-To: <3826b693.3888197@212.225.36.99>
Message-ID: <XFMail.991108204005.mikael@isy.liu.se>

On 08-Nov-99 Endgamer wrote:
 > #####################################
 >  Not a quote, just an on-the-spot observation:
 >  You know you're a geek when your PC costs exactly
 >  15 times as much as your stereo. And gives better
 >  sound quality.<sigh>
 >                                Endgamer
 > #####################################

Sounds like we can define a geek index as 

       Price of PC
  --------------------
  15 * Price of stereo

as an analogue to the audiophile index

        Price of stereo
  --------------------------
  Price of record collection 

off-topic-again-ly y'rs

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    08-Nov-99
Time:    20:33:48

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



From a.eyre at optichrome.com  Fri Nov  5 08:51:08 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Fri, 5 Nov 1999 13:51:08 -0000
Subject: Questions on importing and exec command.
In-Reply-To: <3821E68D.D785F4F3@employees.org>
Message-ID: <000b01bf2794$cfdbb9f0$3acbd9c2@peridot.optichrome.com>

> [snip]
>
> -------- call1.py -------
> for i in range(5,9):
>     print i
>     startit(hold_time=i)
> ---------------------------
>
> [snip]
>
> console > import call1           # importing a code fragment causes
problems
> 5
> exceptions.NameError startit

Although the test2.py script is aware of call1.py, the reverse is not true.
In order to use the 'startit' function in call1.py, it must be in its
namespace.

You can either put:

from test2 import startit

at the top of call1.py, or pass a reference to the function through to the
call1 module.

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com
--------------------------------------------




From strombrg at bingy.acs.uci.edu  Mon Nov 29 23:41:52 1999
From: strombrg at bingy.acs.uci.edu (Dan Stromberg)
Date: 30 Nov 1999 04:41:52 GMT
Subject: difference between invocation of python script on unix and microsoft?
Message-ID: <81vkig$228@news.service.uci.edu>

I wrote a python script for converting "go diagrams" from ascii into a
"structured go format".  If you're into go, you know what I mean, if
not it probably doesn't matter.

I personally use the script under linux.

I made the script available to others via anonymous ftp.  At least one
of the interested users is using a microsoft OS.  This user is having
trouble getting it to work.

Can anyone tell me in what ways using python under microsoft is
different from python under unix?

I assume #! doesn't work.  Assuming I use:

  newstosgf < filename

...under linux, does one need to say:

  python c:/full/path/to/newstosgf < filename

...under microsoft?

(Actually I say just say |newstosgf from within trn on linux - less to
type, but you probably get the idea more from the above than from
this, unless you're a heavy trn user)



From fredrik at pythonware.com  Wed Nov 24 08:17:39 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 24 Nov 1999 14:17:39 +0100
Subject: Python complaints
References: <XFMail.991124113221.mikael@isy.liu.se> <81gj3i$fe3$1@mach.vub.ac.be>
Message-ID: <005201bf367e$4901fef0$f29b12c2@secret.pythonware.com>

Thomas Hamelryck <thamelry at vub.ac.be> wrote:
> I don't want to start the discussion about indentation again; it's no use
> since indenting blocks is here to stay. The point is that if you want to 
> write code that looks ugly or unreadable you will succeed in doing that,
> indentation or not. In many cases, using a variable indentation (depending
> on what construct you are using or the length of a line) leads to code
> that is visuably much more appealing.

you make that sound like a scientific fact.  can
you point us to some research, or is it just yet
another random opinion?

(if it is, we simply disagree.  and for the record,
my opinion is based on many years of experience
from using Python with casual programmers, and
from hacking in huge Python programs written by
other programmers...  what's your opinion based
on?)

</F>




From Gaetan_Corneau at baan.com  Tue Nov 23 11:51:32 1999
From: Gaetan_Corneau at baan.com (Gaetan Corneau)
Date: Tue, 23 Nov 1999 11:51:32 -0500
Subject: Python and COM: a bug?
Message-ID: <816010E2456BD111A48700805FBBE2EEFDF819@ex-quebec-u1.baan.com>

> We have had similar problems here with PythonCOM, and have had to stop
> using it, finding its behaviour too random.

OK, at least I'm not alone. I wish the problem was easy to reproduce :(

______________________________________________________
   Gaetan Corneau
   Software Developer (Quality Assurance Team)
   BaaN  Supply Chain Solutions  
   http://www.baan.com
   E-mail: Gaetan_Corneau at baan.com                  
   Tel: (418) 266-8252          
______________________________________________________
"Profanity is the one language all programmers know best"



 



From joelg at alum.mit.edu  Tue Nov 16 20:20:26 1999
From: joelg at alum.mit.edu (Joel Gould)
Date: Wed, 17 Nov 1999 01:20:26 GMT
Subject: Getting existing instance of Excel
Message-ID: <3832016f.10696885@news.ne.mediaone.net>

Using win32com, you can get a new instance of excel with a command
line:

app=win32com.client.Dispatch('Excel.Application')

But if Excel is already running, is there a way to get a pointer to
the running instance of Excel?

Joel Gould




From chapman at bioreason.com  Tue Nov 16 11:38:11 1999
From: chapman at bioreason.com (Mitch Chapman)
Date: Tue, 16 Nov 1999 09:38:11 -0700
Subject: Line Benchmarks Tkinter vs. wxPython
References: <01bf2fa7$02992ca0$8a42c3d1@Cyberus.cyberus.ca> <slrn8318ec.9dg.bernhard@alpha1.csd.uwm.edu> <01bf2fda$9bf27f40$9c55c3d1@Cyberus.cyberus.ca> <19991116094209.A721671@vislab.epa.gov>
Message-ID: <38318873.CE1B665E@bioreason.com>

Randall Hopper wrote:
> wxWindows fails to build under IRIX (dialog.cpp only knows Sun, Linux, or
> FreeBSD...).  configure --disable-dialupman works around the problem.

Yesterday I also tried compiling wxWindows & wxPython for IRIX.
My situation was complicated by the need to use the O32 ABI together
with the MIPSpro (spel?) compilers.  Although most of the resulting
problems are due to the compiler and not to wxWindows/wxPython,
they may be interesting from a portability point of view.

> wxWindows and wxPython both presume GNU make.  env MAKE=gmake ... works
> around this for wxWindows, but not so for wxPython:
> 
>   > env MAKE=gmake python $WXWIN/utils/wxPython/distrib/build.py -b
>   Makefile created:  Makefile
>   Running: make -f Makefile
>   make: file `Makefile' line 84: Must be a separator (: or ::) for rules (bu39)
>   ...
> 
> Would be nice if build.py took its queues off of os.environ["MAKE"].  I
> hacked sys.MAKE in build.py.

You can also specify the option on the command line, e.g.

build.py -b MAKE=gmake

But this works only when the righthand side of the assignment
does not itself contain an '='.  If the righthand side *does*
contain an '=', you need to copy build.cfg to build.local and 
insert the definition for the variable in question.

Finding out which variables to set is not too hard, but does
require contemplation of the build.py docstring.

> Then it failed finding wx.h.

As you discovered, the fix is to unpack the wxPython tarball
under the utils directory of wxWindows.  And as you noted, the
README is not clear on this point.  It also took me awhile to
realize that you need to define the WXWIN environment variable --
it wasn't needed during the build of wxWindows.


My next set of problems stemmed from the need to use the MIPSpro
compilers w. the O32 ABI.

The most severe of these is that, when compiling for the O32 ABI, 
the C++ compiler does not predefine type bool or the constants true 
or false.

wxWindows handles this situation by providing a typedef for bool,
if one is not already present, and defining the preprocessor macros
TRUE and FALSE.  It would be nice if wxPython used these same
definitions.  It took some time to figure out that I didn't need a
typedef for bool (since it was supplied by wxWindows) but *did* need
definitions for the constants true and false.

The second problem probably also is due to the compiler:
I got error messages for every reference to NULL.  Since the C++
idiom is to use an uncast '0' as NULL, a simple preprocessor
directive (-DNULL=0) solved this problem.

The third problem was the use of C++-style comments in src/libpy.c.
Use of '//' as a comment delimiter in C source code is not
portable.  (Okay, it's portable if everybody's using GCC.  But
it isn't part of the ANSI spec, is it?)

The last problem arose because wxPython expects to use the latest
CVS'd version of SWIG.  We don't have the latest SWIG, and
my account is already overflowing with "personal" builds of
various software packages.  So at this point I gave up.


I did go home last night and install the latest wxGTK and wxPython
packages for Linux.  But that's a subject for a separate post...

-- 
Mitch Chapman
chapman at bioreason.com



From mwh21 at cam.ac.uk  Sun Nov 14 09:25:12 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 14 Nov 1999 14:25:12 +0000
Subject: __str__ vs. __repr__
References: <000001bf275c$02370d20$1d2d153f@tim> <7vvni7$19q0$1@nntp6.u.washington.edu> <382BEC79.EC65913D@maxtal.com.au> <382C6430.8DC07E3E@lemburg.com> <01fd01bf2dce$d2c03cf0$f29b12c2@secret.pythonware.com> <942499879snz@vision25.demon.co.uk>
Message-ID: <m3n1shp2uu.fsf@atrus.jesus.cam.ac.uk>

philh at vision25.demon.co.uk (Phil Hunt) writes:

> In article <01fd01bf2dce$d2c03cf0$f29b12c2 at secret.pythonware.com>
>            fredrik at pythonware.com "Fredrik Lundh" writes:
> > 2) what's the reason byte code cannot be used to recover
> > the original source code
> 
> Perhaps because two different (but obviously functionally
> identical and very similar) source codes could produce the same
> byte code.

I doubt it actually. Python's bytecode is really pretty high
level. Leastaways, I can't think of any examples.

It's one of those things I might get round to writing one of these
days. I don't really know how hard it would be. Not especially, I
suspect.

Regards,
Michael



From frehe491 at student.liu.se  Mon Nov  1 09:26:54 1999
From: frehe491 at student.liu.se (Fredrik Henbjork)
Date: Mon, 1 Nov 1999 15:26:54 +0100
Subject: Destroying files
References: <19991101150451.A26174@optiplex.palga.uucp>
Message-ID: <99110115301000.23977@o112>

On Mon, 01 Nov 1999, you wrote:
> Erno Kuusela wrote:
> > On Thu, 28 Oct 1999 21:38:09 -0400, Gordon McMillan <gmcm at hypernet.com>
> > wrote:
> > 
> > writing zeros over the data does not make it unrecoverable.
> > there are numerous companies that specialise in this kind
> > of recovery.
> 
> You should do it three times.

May I recommend reading this paper:
http://www.cs.auckland.ac.nz/~pgut001/pubs/secure_del.html

--
Fredrik Henbjork

No victim, no crime. Support freedom of mind! 
http://www.maps.org | http://www.norml.org



From mhammond at skippinet.com.au  Wed Nov 24 04:46:48 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 24 Nov 1999 09:46:48 GMT
Subject: Launching Python debugger from embedded Python
References: <11A17AA2B9EAD111BCEA00A0C9B41793034AAF19@molach.origin.ea.com>
Message-ID: <cuO_3.4966$VA6.24292@news-server.bigpond.net.au>

[Posted and mailed]

I would guess the problem is to do with those infernal thread states.  In a
nutshell, unless your code that call back into Python is doing the "right
thing" with the thread-lock, you end up with a NULL thread state.

Its not trivial to explain, but here is my attempt at the short version:

* All of your extension module and object methods must have a
Py_BEGIN_ALLOW_THREADS and Py_END_ALLOW_THREADS pair around the actual call
into your C libraries.

* Any code that calls back into Python must create a CEnterLeavePython class
before doing any Python work.

Mail me for more details...

Mark.

Stidolph, David wrote in message
<11A17AA2B9EAD111BCEA00A0C9B41793034AAF19 at molach.origin.ea.com>...
>I'm getting my company to register for PyWin so we can get the debug
>versions of PyWin, but in the meantime I made a release version of our app
>to test it out.  I have the following code that is called from a Menu
>operation in our own ui:
>
>  def OnMenuDebug(self,x,y,flags):
>    print 'OnMenuDebug'
>    import pywin.debugger
>    try:
>      a = 1 / 0
>    except:
>      pywin.debugger.post_mortem()
>
>In the Output Debug Window of MSDev, I can see that the win32ui.pyd is
>loading, and it loads other DLL's, but it crashes.  I'm going to do another
>release build with all the debug info turned on, but wanted to get this
>question out to see if anyone has ideas to get the debugger working.
>
>OnMenuDebug
>Loaded 'C:\Program Files\Python\Pythonwin\win32ui.pyd', no matching
symbolic
>information found.
>Loaded 'C:\WINNT\system32\PyWinTypes15.dll', no matching symbolic
>information found.
>Loaded 'C:\WINNT\system32\ole32.dll', no matching symbolic information
>found.
>Loaded 'C:\WINNT\system32\oleaut32.dll', no matching symbolic information
>found.
>Loaded 'C:\WINNT\system32\mfc42.dll', no matching symbolic information
>found.
>Loaded 'C:\WINNT\system32\winspool.drv', no matching symbolic information
>found.
>Loaded 'C:\WINNT\system32\shell32.dll', no matching symbolic information
>found.
>Loaded 'C:\WINNT\system32\shlwapi.dll', no matching symbolic information
>found.
>First-chance exception in uo2client.exe (PYTHON15.DLL): 0xC0000005: Access
>Violation.
>The program 'C:\UO2\build-win32\uo2client\Release\uo2client.exe' has exited
>with code 0 (0x0).
>
>
>
>





From aahz at netcom.com  Thu Nov 11 17:02:03 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 11 Nov 1999 22:02:03 GMT
Subject: Beginning programming with Python
References: <sng0ychgwt.fsf@motorola.com>
Message-ID: <80fecr$v62$1@nntp2.atl.mindspring.net>

In article <sng0ychgwt.fsf at motorola.com>,
Aaron Ginn  <aaron.ginn at motorola.com> wrote:
>
>First, I know Perl very well, but I know nothing about Python.  I've
>heard from many sources that Python is the best language to teach
>someone with no programming experience because it emphasizes
>fundamentals and the correct (i.e. best) way to program.  Could anyone
>comment on how the two languages compare in this aspect?  Also, I will
>need to learn Python as well if I choose to go that route.  I am very
>literate with Perl and Tcl/TK.  How quickly can I expect to pick up
>Python?

Python is definitely easier, both from the standpoint of writing code
and from the standpoint of *reading* it.

>Secondly, if I choose to use Python for this, I would like to supply
>each student with a reference book on Python.  I've narrowed it down
>to a choice between the two O'Reilly books, "Programming Python" and
>"Learning Python".  PP apears to have a lot more meat to it, but LP
>appears to be more suited to the beginner.  Assuming the audience
>includes kids that have never programmed before, should I go with LP,
>or will I miss out on anything by not getting the more detailed book?

If you're talking "low-income" and you don't have a fat corporate
partner, I'd strongly suggest that you try the on-line Python docs
first, which includes a tutorial.  Not as good as some of the books (and
I'll add to Ivan's plug because I'm tech-editing his book ;-), but it's
free.  You might get a couple of copies of each of the other books to
fill in the gaps.

Go ahead -- try it for yourself!  (I have at various points in my life
been an expert in BASIC, Pascal, FORTRAN, and Perl (plus a few
proprietary languages), and I like Python best.)
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From guido at cnri.reston.va.us  Tue Nov  9 14:36:50 1999
From: guido at cnri.reston.va.us (Guido van Rossum)
Date: 09 Nov 1999 14:36:50 -0500
Subject: build failure for SunOS 5.5
References: <38269571.14471A7D@lag.ensieg.inpg.fr> <38278EAF.1C6055E2@python.net> <3828062D.B2547E7F@lag.ensieg.inpg.fr>
Message-ID: <5lbt938njh.fsf@eric.cnri.reston.va.us>

Matija Exel <Matija.Exel at lag.ensieg.inpg.fr> writes:

> > Those sound like pretty severe messages in the configure.log.  Did
> > ./configure run to completion without error?
> 
> yes, no errror on ./configure
> 
> > Can you tell which
> > portion of the configure script generates the error messages?
> configure:1388: cc -c -g  conftest.c 1>&5
> "configure", line 1381: invalid source character: '#'
> "configure", line 1381: warning: improper pointer/integer combination: op
> "="
> "configure", line 1381: non-constant initializer: op "NAME"
> "configure", line 1381: invalid source character: '#'
> "configure", line 1381: warning: improper pointer/integer combination: op
> "="
> "configure", line 1381: non-constant initializer: op "NAME"
> "configure", line 1381: syntax error before or at: "() -- "
> "configure", line 1381: warning: syntax error:  empty declaration
> configure: failed program was:
> #line 1377 "configure"
> #include "confdefs.h"
> 
> #define spam(name, doc) {#name, &name, #name "() -- " doc}
> int foo;
> struct {char *name; int *addr; char *doc;} desc = spam(foo, "something");
> 
> int main() {
> ;
> ; return 0; }

Looks like you have a "cc" installed that's not worth the name.
Go find a real C compiler -- you can use GCC (free) or if you insist
on paying, you can buy a real C compiler from Sun...

--Guido van Rossum (home page: http://www.python.org/~guido/)



From none at fake.net  Fri Nov 12 20:29:49 1999
From: none at fake.net (me)
Date: Sat, 13 Nov 1999 01:29:49 GMT
Subject: Win32api.... can it create a Windows shortcut (.lnk) file?
References: <137053459BA7D211B3930090272A62CA8EDD80@newsintern.dspace.de> <zpNW3.53035$Rx2.650681@news1.rdc1.bc.home.com>
Message-ID: <ha3X3.55179$Rx2.678471@news1.rdc1.bc.home.com>

Doh!

I guess with all the time I was spending reading the docs for win32com, I
missed the typo I had in the target path name...

Oh well, you live and you learn, and there's stupider mistakes you can make.

It works great, btw... thanks again for all the responses.

Kevin.




me <none at fake.net> wrote in message
news:zpNW3.53035$Rx2.650681 at news1.rdc1.bc.home.com...
> That's a nice clean example.... but I'm having one small problem, even
> taking it exactly as you put it:
>
>
> ##########################################
> Traceback (innermost last):
>   File "C:\Program
> Files\Python\Win32\Pythonwin\pywin\framework\scriptutils.py", line 237, in
> RunScript
>     exec codeObject in __main__.__dict__
>   File "C:\PROGRA~1\PYTHON\MYPROJ~1\WINTOOLS\wintools.py", line 144, in ?
>     CreateShortCut(Path,Target,Arguments,StartIn,Icon,Description)
>   File "C:\PROGRA~1\PYTHON\MYPROJ~1\WINTOOLS\wintools.py", line 132, in
> CreateShortCut
>     persist.Save(Path, 1)
> com_error: (-2147467259, 'Unspecified error', None, None)
> ##########################################
>
> Unfortunately, "Unspecified error" doesn't help me too much... q:].  I'll
be
> looking into that tonite, but if you have a quick fix, I'd appreciate it.
> I've been using the win32api module a lot, and it's great, but not
pythoncom
> a lot yet...
>
> q:]
>
> Kevin.
>






From jonkers at gmx.net  Fri Nov  5 16:19:27 1999
From: jonkers at gmx.net (Sander Jonkers - GMX)
Date: Fri, 5 Nov 1999 22:19:27 +0100
Subject: rs232
References: <3822A8D8.8093F56D@obop.com.pl>
Message-ID: <ebLVqK9J$GA.213@net003s>

Przemys?aw G. Gawro?ski <gawron at obop.com.pl> wrote in message
news:3822A8D8.8093F56D at obop.com.pl...
> Hello, I'm very new to Python but I'm having a great time learning it !
>
> What is the simplest way to read from the RS232 port under windows,
> maybe there is a module for that ready ?

A part of learning Python (and probably anything else) is reading the FAQ.
http://www.python.org/doc/FAQ.html There you will find the FAQ "4.68. How do
I access the serial (RS232) port?"

Sander





From alex at somewhere.round.here  Tue Nov  2 13:41:25 1999
From: alex at somewhere.round.here (Alex)
Date: 02 Nov 1999 13:41:25 -0500
Subject: Printing strings
References: <381F00E1.721BB47F@cornell.edu>
Message-ID: <etdiu3krb1m.fsf@w20-575-100.mit.edu>

> I'm trying to do something like the following:
> 
> a=4.0
> b=2.0
> print '2(%3.1f)' % a,b
> 
> where I really mean
> 
> print '%3.1f %3.1f' % a,b

You could use

print string.join(2*['%3.1f'])%(a,b)

I suppose.

Alex.



From vnesek at maja.zesoi.fer.hr  Thu Nov  4 18:30:59 1999
From: vnesek at maja.zesoi.fer.hr (Vjekoslav NESEK)
Date: 4 Nov 1999 23:30:59 GMT
Subject: JavaCard and JPython?
References: <19991103.11420500@kahua.transpacific.com.hk>
Message-ID: <slrn8245og.pkv.vnesek@maja.zesoi.fer.hr>

In article <19991103.11420500 at kahua.transpacific.com.hk>, 
Scott Deerwester wrote:
>Is there any thought of implementing the JavaCard API using Jpython? 
>The prospect of being able to program Smartcards in Python is very 
>appealing indeed!

No since JavaCard API is running on SmartCards. Say up to 32K for
your .class and few hundred bytes for memory. 8-16 bit processors
(I know there are 32 bit prototypes). It is like pushing an elephant
into a paper bag :)

-- 
// Vjeko (vnesek at zesoi.fer.hr)

import java.advertising.*;
new URL("http://diana.zesoi.fer.hr/~vnesek/");



From ionel at psy.uva.nl  Wed Nov 24 13:49:25 1999
From: ionel at psy.uva.nl (Ionel Simionescu)
Date: Wed, 24 Nov 1999 19:49:25 +0100
Subject: whrandom seed question
References: <81eu22$411al$1@umbc7.umbc.edu>
Message-ID: <81hbsg$5nh@mail.psy.uva.nl>

| Am I missing something???
|

If I recall correctly, by convention, seeding with zeros is equivalent to
use the seed automatically derived from the time value.

ionel







From wtanksle at hawking.armored.net  Mon Nov 15 17:34:00 1999
From: wtanksle at hawking.armored.net (William Tanksley)
Date: 15 Nov 1999 22:34:00 GMT
Subject: What are Aspects? (Re: How can I make my assertions smarter?)
References: <Pine.WNT.4.05.9911032042440.109-100000@david.ski.org> <slrn82h5o6.26j.wtanksle@hawking.armored.net> <80eocf$8rc$1@nntp5.atl.mindspring.net> <WIDW3.331$Ae3.219038@news-dal.corridex.com> <80fdcc$d4f$1@nntp6.atl.mindspring.net>
Message-ID: <slrn8312kc.tv9.wtanksle@hawking.armored.net>

On 11 Nov 1999 21:44:44 GMT, Aahz Maruch wrote:
>Evan Simpson <evan at tokenexchange.com> wrote:

>>I think that this depends on the language in which you write your Aspects (I
>>don't mean C/Python/etc, I mean the special Aspect-specification language).
>>In the examples I looked at, pattern-matching on the class, method, and
>>method arguments determined where each part of an Aspect would be applied.

>So it's essentially a kind of pre-processor?

That's an excellent way of thinking about it.  It's essentially a semantic
preprocessor.  Instead of searching for and inserting text, like the C
preprocessor, it searches for and replaces behavior in certain contexts.

Unlike the C preprocessor, though, most aspect-oriented programming
systems don't introduce new symbols to the language.

>                      --- Aahz (@netcom.com)

-- 
-William "Billy" Tanksley



From wuhy at tcts.seed.net.tw  Fri Nov 12 17:58:54 1999
From: wuhy at tcts.seed.net.tw (Mark Wu)
Date: 12 Nov 1999 22:58:54 GMT
Subject: [HELP] TKinter could not run in Python!
Message-ID: <80i63e$7t4$1@news.seed.net.tw>

Dear Sir:

    This is my first time to use Python, I tried to active the Tkinter, but
the interpreter always tell me:

error message:
----------------------------------------------------------------------------
--------------------

  File "<stdin>", line 1, in ?
  File "D:\Python\Lib\lib-tk\Tkinter.py", line 1947, in _test
    root = Tk()
  File "D:\Python\Lib\lib-tk\Tkinter.py", line 886, in __init__
    self.tk = _tkinter.create(screenName, baseName, className)
TclError: Can't find a usable init.tcl in the following directories:
    {d:\progra~1\tcl\lib\tcl8.1} {} ./lib/tcl8.0 D:/Python/tcl8.0/library
D:/Pyt
hon/Tools/library

d:/progra~1/tcl/lib/tcl8.1/init.tcl: version conflict for package "Tcl":
have 8.
0, need 8.1
version conflict for package "Tcl": have 8.0, need 8.1
    while executing
"package require -exact Tcl 8.1"
    (file "d:/progra~1/tcl/lib/tcl8.1/init.tcl" line 19)
    invoked from within
"source d:/progra~1/tcl/lib/tcl8.1/init.tcl"
    ("uplevel" body line 1)
    invoked from within
"uplevel #0 [list source $tclfile]"


This probably means that Tcl wasn't installed properly.
----------------------------------------------------------------------------
--------------------

My system is window NT, with TCL/TK 8.1 (already installled before install
python)...

Any suggestion and help is very appreciate :)

Best Regards, Mark





From neelk at brick.cswv.com  Tue Nov 30 20:56:47 1999
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 1 Dec 1999 01:56:47 GMT
Subject: predicate classes
References: <81uo0t$2rr$1@nnrp1.deja.com>
Message-ID: <slrn8492gk.4a8.neelk@brick.cswv.com>

Justus Pendleton <justus at my-deja.com> wrote:
>I decided to broaden my horizons so I was reading about/playing around with
>Cecil.	One of the things Cecil has that I hadn't heard of before are
>predicate classes:
>
>(http://www.cs.washington.edu/research/projects/cecil/cecil/www/www/Papers/
> predicate-classes.html), which I thought were nifty beyond all belief.

Good for you! I like Cecil a lot, too. You might also be interested in
Self (http://self.sunlabs.com) -- it implements dynamic inheritance,
which is one of the motivations for Chambers' invention of predicate
classes.

>I also thought it would be nifty if Python had them so I got the source
>from the CVS respository and started hacking the grammar and realized I
>had no idea what I was doing :-).
>
>Before getting in any deeper in something I only vaguely understand I
>thought I would ask the opinions of people who know more about Python
>than I whether they think it would be easy/hard/well-damn-nigh-impossible 
>to add this kind of dynamic retyping of objects based on predicates.

Fortunately, you don't need to hack the Python source in order to
experiment with dynamic inheritance or predicate classes in Python,
because you can change a class's inheritance tree by assigning to a
class's __bases__ attribute at runtime.

Here's a quick, rather kludged-together example:

class Foo:
    x = 4

class Bar:
    predicate = 1
    __bases__ = (Foo,)
    #
    def __setattr__(self, name, value):
        if name == 'predicate':
            self.__class__.predicate = value
            if value:
                self.__class__.__bases__ = (Foo,)
            else:
                self.__class__.__bases__ = ()

At the interpreter prompt:

>>> s = Bar()
>>> isinstance(s, Foo)
1
>>> s.x
4
>>> s.predicate = 0
>>> isinstance(s, Foo)
0
>>> s.x
Traceback (innermost last):
  File "<stdin>", line 1, in ?
AttributeError: x

Whether or not s, the instance of Bar, inherited from Foo depends 
on what the value of Bar.predicate was. Note that this isn't the 
same thing  as real predicate dispatch, yet, since different 
instances of Foo can't have different virtual classes. 

To make that work, you would need to create a custom sequence object
for your __bases__ object and hack together a metaclass to install it
cleanly. This isn't a simple project, but fortunately, it's not a big
job either -- most of the work is conceptual rather than writing reams
of code.

It will also be lots of fun; my current Python project right now is
writing a multimethod class for Python and I can vouch for the hack
value of something like this.

Some day in the future, *after* you have a good idea of what semantics
you want to implement, you can create a C extension type implementing
predicate classes and then extend the interpreter. But that's for
later; first prototype it in Python itself before wasting time writing
C code. Remember you won't know what the right thing is until you've
used your metaclass in a variety of projects, and you will feel much
easier about modifying and rewriting a Python class than a C
extension.

>(I did a quick search on deja and didn't find any talk about this, so
>apologies in advance if this is something that has been as talked to
>death as the indentation thing.)

Not at all! Cecil, Dylan and Self are all great languages to use to
discover Pythonic features in. :)


Neel



From tim_one at email.msn.com  Fri Nov  5 02:04:30 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Fri, 5 Nov 1999 02:04:30 -0500
Subject: __str__ vs. __repr__
In-Reply-To: <199911041325.IAA14001@eric.cnri.reston.va.us>
Message-ID: <000001bf275c$02370d20$1d2d153f@tim>

[amazingly enough, still sticking to what the subject line sez <ahem>!]

Let me back off to what repr and str "should do":

repr(obj) should return a string such that
    eval(repr(obj)) == obj
I'd go further and formalize a property that's currently honored but
implicitly:  the string returned by repr should contain only characters c
such that
    c in "\t\n" or 32 <= ord(c) < 127
i.e. it should be restricted to the set of 7-bit ASCII characters that the C
std guarantees can be read back in faithfully in text mode.  Together, those
properties spell out what repr() is "good for":  a highly portable and
eval'able text representation that captures *everything* important about an
object's value.

In contrast, str() should return a string that's pleasant for people to
read.  This may involve suppressing obscure or minor details, displaying it
in a form that's inefficient-- or impossible --to process, or even plain
lying.  Examples include the Rat class I used before (the ratio of two huge
integers is exact but unhelpful), a DateTime class that stores times as
seconds from the epoch (ditto), or a Matrix class whose instances may
contain megabytes of information (and so a faithful repr() string may be
intolerable almost all the time).

Viewed that way, I'd collapse all of your (Guido's) characterizations of
what I'm after into one:  repr should never be invoked implicitly.  If you
want a faithful (but often unhelpful) string, use repr(obj) or `obj`
explicitly.  Part of not invoking repr implicitly includes containers not
inventing repr() calls out of thin air <wink>.

David Ascher suggested that str(long) drop the trailing "L", and that's a
good example.  The "L" is appropriate for repr(), for so long as Python
maintains such a sharp distinction between ints and longs, but is really of
no help to most users most of the time.

repr() currently cheats in another way:  for purely aesthetic reasons,
repr(float) and repr(complex) don't generate enough digits to allow exact
reconstruction of their arguments (assuming high-quality float<->string in
the platform libc).  This is a case where an argument appropriate to str()
was applied to repr(), not because it makes *sense* for repr, but because so
much output goes thru an *implicit* repr() now and you didn't want to see
all those "ugly" long float strings <0.50000000000000079 wink>.

repr(float) should do its job correctly here; str(float) should continue to
produce the "nice" (but numerically inadequate) strings.  (BTW, on an
IEEE-754 platform repr(float) should generate up to 17 significant digits
(suppression of trailing zeroes is fine) -- the 754 std guarantees that if
conforming output generates that many, a conforming input routine will
exactly reconstruct the original value -- note that this does *not* require
best-possible conversion in either direction -- it turns out that 17 is
enough for somewhat sloppy conversions to succeed, and many platforms are up
to this easier task)

Onward.  Or backward:

[Tim]
>> In a world where containers passed str() down, a container's
>> str() would presumably be responsible for adding disambiguating
>> delimeters to element str() results when needed (the container
>> knows its own output syntax, and can examine the strings produced
>> by its elements ...

[Guido]
> Hm...  What kind of things would you expect e.g. the list str() to do
> to its item str()s?  Put backslashes before commas?

This depends on how seriously you take all this <wink>.

If you take it very seriously, __str__ should take a second flag argument,
saying whether the consumer of the string will be embedding the string in a
larger context or displaying it directly.  Then it's up to __str__ to put
its own brand of delimiters around its output when appropriate.

I suspect that's overkill, though -- that the problem here is very specific
to a string object's __str__ (containers will put matching brackets at each
end of their str() or repr() output regardless, numbers of various sort
don't need delimiters, and a user class __str__ generally produces a highly
stylized string -- seems that it's *only* string objects that have wildly
unpredictable display forms).

If true, it may work fine just to be simple and pragmatic:  let containers
special-case the snot out of elements of string type, sticking a pair of
quotes around what string's __str__ returns and backslash-escaping any
embedded quotes of the same type.

Then

>>> names = ["Fran?ois", "Tim", "Gu'ido"]
>>> names   # same as str(names) in this make-believe world
['Fran?ois', 'Tim', 'Gu\'ido']
    or
['Fran?ois', 'Tim', "Gu'ido"]

is what I'd expect.

I certainly don't want to get rid of the commas, colons, parens, braces and
brackets for list, tuple and dict str() output:  they're very helpful!  I
just want to be able to read what's *between* all that stuff.

> These are all good points.

Ya, well ... everything sounds kinda good before it's implemented <0.6
wink>.

> In a typical scenario which I wanted to avoid, a user has a variable
> containing the string '1' but mistakenly believes that it contains the
> integer 1.  (This happens a lot, e.g. it could be read from a file
> containing numbers.)  The user tries various numeric operations on the
> variable and they all raise exceptions.  The user is inexperienced and
> doesn't understand what the exceptions are, but gets the idea to
> display its value to see if something's wrong with it.  One of the
> first things users learn is to use interactive Python as a power
> calculator, so my hypothetical user just types the name of the
> variable.  If this would use str() to format the value, the user is no
> wiser, and perhaps more confused, since str('1') is the same as
> str(1).

Understood and appreciated.  Using the pragmatic approach above, I would
have PRINT_EXPR call the same routine that container str() implementations
call, i.e. one that special-cases the snot out of strings.  Then

>>> x, y = 1, '1'
>>> x
1
>>> y
'1'
>>> print x
1
>>> print y
1
>>>

is what I'd expect.

> ...
> We can then argue over what str() of a list L should return; one
> extreme possibility would be to return string.join(map(str, L));

I definitely don't want to lose the brackets or the commas.

> a slightly less radical solution would be
>     '[' + string.join(map(str, L), ', ') + ']'

Yes, but replacing str with str_special_casing_the_snot_out_of_strings.

> In the first case, your last example would go like this:
>
> >>> names
> Fran?ois Tim
> >>>

Yuck!

> while the [second] choice would give
>
> >>> names
> [Fran?ois, Tim]
> >>>

As above, ['Fran?ois', 'Tim'] is my ideal.

> There may be other solutions -- e.g. in Tcl, a list is displayed as
> the items separated by spaces, with the proviso that items containing
> spaces are displayed inside (matching) curly braces; unmatched braces
> are displayed using backslashes, guaranteeing that the output can be
> parsed back into a list with the same value as the original.  (Hey!
> That's the same as Python's rule!  So why does it work in Tcl?
> Because variables only contain strings, and the equivalent of the Rat
> class used above can't be coded in Tcl.)

I'm explicitly giving up the property that str(obj) always be eval'able in a
sensible way.  That's repr()'s job.  For example, I expect

>>> x = "\\t"
>>> x
'\t'
>>>

and eval('\t') certainly isn't x.  This is a tradeoff:  we stop pretending
that str() is adequate for equality under evalability <wink>, but take EUE
seriously for repr().  In return, str() gets to do what it was meant to do:
produce "nice" strings, without compromise for the sake of faux
(inconsistent, unpredictable) EUE.

> The problem with displaying 'Fran?ois' has been mentioned to me
> before.  (By the way, I have no idea how to *type* that.  I'm just
> cutting and pasting it from Tim's message.)

European keyboard, programmable keyboard, C-q something-or-other in Emacs,
Alt+0231 in Windows (keyboard generation of any 8-bit code is built in to
Windows keyboard drivers), or even copy+paste from Accessories->Character
Map under Windows.  My employer spends a lot of time worrying about other
countries' foolish character sets <wink>.

> There's another scenario I was trying to avoid.  This is probably
> something that happened once too many times when I was young and
> innocent, so I may be overracting.  Consider the following:
>
> >>> f = open("somefile")
> >>> a = f.readline()
> >>> print a
> %????1.3
>
> >>>
>
> Now this example is relatively harmless.  Using repr(), I see that the
> string contains a \r character that caused the cursor to back up to
> the start of the line, overwriting what was already written:
>
> >>> a
> '%PDF-1.3\015%\342\343\317\323\015\012'
> >>>
>
> But the thing that some big bad program did to me long ago was more
> like spit out several thousand garbage bytes which contained enough
> escape sequences to lock up my terminal requiring me to powercycle and
> log in again.  (The fact that the story refers to a terminal indicates
> how long ago this was. :-)
>
> So I vowed that *my* language would not (easily) let this happen by
> accident, and the way I enforced that was by making sure that all
> non-ASCII characters would be printed as octal escapes, unless you use
> the print statement.

An irony is that I've locked up a DOS box doing this on many occasions -- in
a script, "print a" is not exactly a *hard* mistake to fall into <wink>.

I'm not unsympathetic, but this is something you really can't stop without
rendering Python useless to grownups.  How does Python's Unicode story
relate to this?  Displaying a Unicode string as a pile of \uxxxx (whatever)
escapes defeats the whole purpose of Unicode.  Ditto displaying it in UTF-7.
OTOH, for some number of years to come it will be a rare display device that
*won't* treat Unicode 16-bit values (whether or not UTF-8 encoded) as a pile
of 8-bit control codes.

> It's a separate dilemma from the other examples.  My problem here is
> that I hate to make assumptions about the character set in use.  How
> do I know that '\237' is unprintable but '\241' is a printable
> character?

We have no idea.

> How do I know that the latter is an upside-down exclamation point?

Ditto.

> Should I really assume stdout is capable of displaying Latin-1?

No.  But I don't grasp why you think you need to know *anything* about this.
Until Unicode takes over the world, there's nothing you can do other than
tell users the truth:  "most of" printable 7-bit ASCII displays the same way
across the world, but outside of that *all* bets are off.  It will vary by
all of OS, display device, displaying program, and user configuration
choices.

> Strictly, the str() function doesn't even know that it's output is
> going to stdout.

That's right, and C doesn't even guarantee that I can read Fran?ois back in
from a text file!  It's that bad.  *So* bad that everyone lives with it and
never notices it <0.9 wink>.

> I suppose I could use isprint(), and then Fran?ois could use the locale
> module in his $PYTHONSTARTUP file to make it do the right thing.  Is that
> good enough?

I don't know what you're trying to *accomplish* here.  isprint() is a piece
of crap, not least because what can or can't be displayed has much more to
do with the font you're currently using than with anything C knows about.
That is, it's outside C's areas of competence.

For example, here's the relatively new euro symbol:  ?.  That's not in
Latin-1.  Windows added it as an extension to Latin-1, at hex code 0x80 (one
of the 32 codes Latin-1 didn't assign glyphs to).  In my mailer at the
moment, even on Windows, it shows up as a square box.  That's because my
font at the moment happens to be Courier.  If I switch my font to Courier
New, it shows up as intended -- but even then only because I happened to
install the service pack that added the euro symbol, and Courier New (unlike
Courier) is a "code page 1252" font.  On any non-Windows system, God only
knows what you'll see.  The point is that C's isprint(0x80) can't give a
reasonable answer even sticking solely to my machine!  What I can or can't
display has, alas, nothing to do with C.

> (I just tried this briefly.  It seems that somehow the locale module
> doesn't affect this?!?!

Locale is supposed to affect all of C's isxxx functions, but I wouldn't bet
that most implementations do it correctly.  As above, even if they did,
what's actually displayable is a different issue.

> I still think that ['a', 'b'] should be displayed like that, and not
> like [a, b].

Me too!  I don't want to make the currently-readable unreadable -- I want to
keep that and get the converse too.

> I'm not sure what that to do about a dict of Rat() items,

str(recip) (my example) produces {0.100: 1.000, 1.000: 0.100} (which follows
automatically from str(dict) "passing str() down", and that neither the keys
nor the values are themselves of string type); repr(recip) produces the
unreadable but precise and eval'able string that's produced today (which
follows automatically from repr(dict) passing repr() down).

> except perhaps giving up on the fiction that __repr__() should
> return something expression-like for user-defined classes...
>
> Any suggestions?

Yes:  stop being so radical <wink>.  That eval(repr(obj)) == obj is a
*wonderful* property of the language design!  I don't want to lose that
either.  My claim is that repr() is rarely appropriate, but is essential
when it is.  This is very often very apparent in user-defined classes (at
least mine <wink -- but I seem to hit this every time!>).  It was hard to
see at first because the builtin types other than string treated repr() and
str() as synonyms, and nobody bitched enough about str(long)'s trailing "L",
and I didn't bitch enough about repr(float)'s == str(float)'s systematic
inaccuracy, and Python's European groupies didn't bitch enough about
repr(euro_string) being an unreadable backslashed mess.

IOW, I don't want to get rid of the str/repr design!  I want to take it
seriously and do all the obvious <wink> things that follow from that.

telling-someone-what-they-really-think-takes-many-words-ly y'rs  - tim





From fredrik at pythonware.com  Mon Nov 29 09:24:51 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 29 Nov 1999 15:24:51 +0100
Subject: Unbuffered stdout
References: <81u1k2$27qkk$1@oceanite.cybercable.fr>
Message-ID: <046a01bf3a75$80b84d90$f29b12c2@secret.pythonware.com>

leon <leonar_n at epita.fr> wrote:
> I'm developping an IRC like software which must work on the Web. I can't use
> Java, and I'm using Python to program the CGI. But, I have a big problem : I
> don't know how to make the stdout unbuffered.
> I tried to make some
>     sys.stdout.write("test")
>     sys.stdout.flush()
> but it don't work.

looks fine to me.  maybe something else
is wrong?

anyway, you could try running the inter-
preter with the -u option:

$ python -?
Unknown option: -?
usage: python [option] ... [-c cmd | file | -] [arg] ...
Options and arguments (and corresponding environment variables):
...
-u     : unbuffered binary stdout and stderr (also PYTHONUNBUFFERED=x)
...

</F>




From parkw at better.net  Thu Nov 25 17:45:44 1999
From: parkw at better.net (William Park)
Date: Thu, 25 Nov 1999 17:45:44 -0500
Subject: Journaling Code
In-Reply-To: <81k6m9$gkm$1@newsie2.cent.net>
References: <81k6m9$gkm$1@newsie2.cent.net>
Message-ID: <19991125174544.A1164@better.net>

On Thu, Nov 25, 1999 at 03:36:45PM -0500, Norman Vine wrote:
> I am involved with a new Open Source 3D Modeller project
> and we are investigating various undo-redo mechanism's.
> 
> A member of the group suggested implementing a scripting
> layer between the modeller and the GUI, and journalling all
> user invoked scripts in order to be able to rebuild the model
> from a saved state by simply replaying the journaled code.
> 
> I am wondering if anyone has used Python for anything similar,
> and if so if they would be willing to share their knowledge.
> 
> Cheers
> 
> Norman Vine
> 
> archives  http://www.topica.com/lists/3dmodeller
> URL ; http://www.woodsoup.org/~sbaker/modellers

Not coming from 3D rendering background... but, wouldn't it be easier
to have your modeller produce "log" file of anything coming in?



From moshez at math.huji.ac.il  Thu Nov 11 02:52:34 1999
From: moshez at math.huji.ac.il (Moshe Zadka)
Date: Thu, 11 Nov 1999 09:52:34 +0200 (IST)
Subject: How to perform syntax check in Python
In-Reply-To: <382A2384.CC37FBA8@hp.com>
Message-ID: <Pine.SOL.3.96.991111094750.5663A-100000@sundial>

On Wed, 10 Nov 1999, Meera wrote:

> Hello everybody,
> 
> I am looking for  a utility to perform syntax validation in python.
> Consider following code -
<and really wants some way to detect undeclared variables>
> When we have multi-developer and multi-platform environment - such
> feature is crucial.  

Ummm....what planet are you from?

Python has no notion of declaring variables at compile time, only at
run-time. That means program flow can impact what name are "alive". This
means Python has to solve halting-problem-equivalent problems in order to
check for undeclared variables. I don't know what state HP is in, but if
you know how to solve the halting problem, you should patent it, and build
a brand new company. Guido has chosen not to release that technology into
Python for the next 5 years.

--
Moshe Zadka <mzadka at geocities.com>. 
INTERNET: Learn what you know.
Share what you don't.




From pereira at research.att.com  Wed Nov  3 23:47:05 1999
From: pereira at research.att.com (Fernando Pereira)
Date: Thu, 4 Nov 1999 04:47:05 GMT
Subject: A Comparison of Python and Ruby Extension Interfaces (Re: Comparison between Python and "Ruby")
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp> <7vjtrc$hmm$1@newshost.accu.uu.nl> <38204614.2EAC6051@compaq.com> <021119992327527678%pereira@research.att.com> <38215CE7.70C3D560@compaq.com>
Message-ID: <031119992347056096%pereira@research.att.com>

In article <38215CE7.70C3D560 at compaq.com>, Greg Ewing
<greg.ewing at compaq.com> wrote:

> Fernando Pereira wrote:
> > 
> > What's the effect of this on portability? For instance, how are
> > pointers to live objects from the C stack handled? As is well-known,
> > this cannot be done portably.
> 
> It scans the whole C stack and marks anything that looks like
> it might be a pointer to a Ruby object. It assumes that setjmp()
> saves any registers that might contain such pointers. There are
> a few special cases coded in. Porting to a new system might
> require this piece of code to be adjusted, although it looks
> to me like it should work under most sensible architectures.

In my experience, there's a big difference between what you and I call
"sensible" and what compiler and OS developers try to get away with,
not to mention the vagaries of particular hardware. For example, what
is the state of the stack, registers, etc. when a Unix signal handler
is called? What if that handler (maybe by calling back into the
higher-level language) forces a GC?
> 
> Global variables in C extensions which contain Ruby pointers
> have to be explicitly made known during initialisation.
> 
> > > * Ruby seems to be able to raise exceptions without
> > > having to explicitly propagate them back out of routines,
> 
> > Same comment. There is no portable way of unwinding the C stack. What
> > does that mean for Ruby's portability?
> 
> It uses setjmp/longjmp. Presumably these work correctly on
> any system where they are implemented. Porting Ruby to a
> system where they're not implemented could be difficult.

Well, they are typically implemented, but they may not work as one
expects (eg. call longjmp from a signal handler). 

The reason I picked on both issues is that they have caused
never-ending problems for language implementations that want to coexist
flexibly with C and the OS. I've come to appreciate the choices made in
Python after spending many late, frustrating nights tracking down
obscure bugs in Prolog implementations, GNU Emacs ports (yes, GNU Emacs
*is* a language), Scheme scripting layers, and having talked often with
language implementers for other languages (eg. Lisp, ML) about how
tricky it is to get these things right in language implementations that
make too many assumptions about "sensible architecture", just to have
the next release of the C compiler or OS break things again.

-- F



From hnowak at cuci.nl  Sun Nov 14 05:19:41 1999
From: hnowak at cuci.nl (Hans Nowak)
Date: Sun, 14 Nov 1999 11:19:41 +0100
Subject: itoa on python?
In-Reply-To: <382e77d5.26188609@news.remarq.com>
Message-ID: <199911141021.LAA06164@dionysus.fw.cuci.nl>

On 14 Nov 99, at 8:52, Gusdvg wrote:

> How can I change an integer (or any other type of number) to a string? In
> C I used itoa but I couldn't find it in the Python documentation.

Try one of these:

x = 42

print str(x)
print `x`
print "%d" % x

The first two work with other types of numbers too. The third one is 
like C's *printf, and is useful if you need formatting and such.

Blessed be,

--Hans Nowak (zephyrfalcon at hvision.nl)
Homepage: http://fly.to/zephyrfalcon
You call me a masterless man. You are wrong. I am my own master.



From conversy at lri.fr  Fri Nov 12 09:46:45 1999
From: conversy at lri.fr (Stephane Conversy)
Date: Fri, 12 Nov 1999 15:46:45 +0100
Subject: Tkinter and after
Message-ID: <382C2854.E9494EFE@lri.fr>

Hello.

It is said in Tkinter doc that after() is like a call to time.sleep()
wich blocks
Tkinter.

there is a _tkinter.createtimerhandler(), but the returned objects can
not be
cancelled (as it should be with after_cancel).

So how to do an after() with  Tkinter ? I implemented a threaded timer,
but
I need to do things in the main thread, not in the timer thread.

and I didn't find it in the faq...

--
St?phane Conversy
http://www-ihm.lri.fr/~conversy/
mailto:conversy at lri.fr






From mhammond at skippinet.com.au  Wed Nov 24 08:35:10 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 24 Nov 1999 13:35:10 GMT
Subject: Upgrading Python on Windows
References: <383BCBCA.5804D43E@python.net>
Message-ID: <iQR_3.5212$VA6.25796@news-server.bigpond.net.au>

Thomas A. Bryan wrote in message <383BCBCA.5804D43E at python.net>...

>I-can-troubleshoot-a-UNIX-install,-but-Windows-scares-me-ly yours

You should be able to uninstall Python 1.5.1 from Control Panel->Add Remove
Programs.  Id suggest doing before installing the new version.

If you have no other apps that need Tcl/Tk, then I would suggest
uninstalling them too.

Nothing-to-be-scared-of-ly,

Mark.






From kern at caltech.edu  Sat Nov  6 04:02:48 1999
From: kern at caltech.edu (Robert Kern)
Date: 6 Nov 1999 09:02:48 GMT
Subject: Problems Building Python under Cygwin
References: <7vtath$n1j$1@nnrp1.deja.com>
Message-ID: <800qro$k95@gap.cco.caltech.edu>

In article <7vtath$n1j$1 at nnrp1.deja.com>, salvadorej at my-deja.com writes:
> I'm having some problems building Python under Cygwin on Windows NT 4.
> 
> 1) the first problem I had was relatively easy to deal with. Towards
> the end of the make process, there's a command to copy the python
> executable from the Modules directory. The problem is, the makefile
> expects to copy something called 'python', but the executable produced
> was 'python.exe' (presumably because I'm under Windows). So cp can't
> complete this command, and the makefile chokes. Fortunately, at this
> point, the make process is pretty much done, so all I had to do was go
> into the Modules directory, and copy python.exe to the directory above,
> and it worked. SO: did I mess up? Or is this something wrong with the
> makefile/configure setup that needs to be fixed?

There is a line in some of the Makefile's:

EXE=

Changing that to:

EXE=.exe

is what you should only need to do.  It's been a while since I did that, 
though, and I seem to remember that it did not solve all of the problems.
Anyways, the problem is probably moot; see below.

> 2) the second problem I haven't solved. What I really wanted Python for
> was to build Zope from scratch. 

Do you need to compile Zope statically into Python, or were you simply
constrained by the lack of MSVC?  If the latter, you do not need to compile
Python and Zope with Cygwin.  You can use mingw32 (using Mumit Khan's native
build, not the --mno-cygwin switch on Cygwin's compiler) to compile Zope's
extension modules and use them with the standard Win32 Python interpreter. 
I did it for one of the Zope 2 betas.

Se my page:
http://starship.python.net/crew/kernr/mingw32/Notes.html

for details.


--
Robert Kern
kern at caltech.edu

"In the fields of hell where the grass grows high
 are the graves of dreams allowed to die."
       -- Richard Harter



From gerrit.holl at pobox.com  Mon Nov 29 11:43:22 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Mon, 29 Nov 1999 17:43:22 +0100
Subject: Unbuffered stdout
In-Reply-To: <81u2gm$26v0f$1@oceanite.cybercable.fr>; from leonar_n@epita.fr on Mon, Nov 29, 1999 at 03:32:25PM +0100
References: <81u1k2$27qkk$1@oceanite.cybercable.fr> <046a01bf3a75$80b84d90$f29b12c2@secret.pythonware.com> <81u2gm$26v0f$1@oceanite.cybercable.fr>
Message-ID: <19991129174322.A4550@stopcontact.palga.uucp>

leon wrote:
> I use it into a cgi script, with apache, and for my IRC like, I must do
> a thing like :
>     while 1:
>         strLine = self.pRoomFile.readline()
>         while strLine:
>             sys.stdout.write(strLine)
>             sys.stdout.flush()
>         strLine = self.pRoomFile.readline()
>         time.sleep(1)

Are you sure you printed the 'Content-type: text/html\n\n'? And are you
also sure you checked the error logs? Does it run from the command line?

regards,
Gerrit.

-- 
"However, complexity is not always the enemy."

  -- Larry Wall (Open Sources, 1999 O'Reilly and Associates)
  5:42pm  up  2:26, 16 users,  load average: 1.29, 1.29, 1.01



From g_will at cyberus.ca  Sun Nov 28 11:29:35 1999
From: g_will at cyberus.ca (Gordon Williams)
Date: Sun, 28 Nov 1999 16:29:35 GMT
Subject: Pythonwin Reset
References: <01bf38df$8127e0c0$3a42c3d1@Cyberus.cyberus.ca> <_X004.7432$VA6.38559@news-server.bigpond.net.au>
Message-ID: <01bf39bd$a28831a0$4155c3d1@Cyberus.cyberus.ca>

Hi Mark,

When I try it I get:

>>> __main__.__dict__.clear()
Traceback (innermost last):
  File "<interactive input>", line 0, in ?
NameError: __main__

what has gone wrong?

Regards,

Gordon Williams

Mark Hammond <mhammond at skippinet.com.au> wrote in article
<_X004.7432$VA6.38559 at news-server.bigpond.net.au>...
> Nope.  The closes Pythonwin could really do is to perform a
> __main__.__dict__.clear(), and that is pretty easy to type.  There is no
way
> Pythonwin could perform a more intrusive reset, as Pythonwin is written
in
> Python!
> 
> Mark.
> 
> Gordon Williams wrote in message
> <01bf38df$8127e0c0$3a42c3d1 at Cyberus.cyberus.ca>...
> >Hello All,
> >
> >Is there a way to reset Pythonwin IDE between runs.  I want to eliminate
> >any rogue objects or modules that are still in memory from the previous
run
> >of the program.
> >
> >Thanks
> >
> >Gordon Williams
> 
> 
> 



From g_will at cyberus.ca  Mon Nov 15 18:04:19 1999
From: g_will at cyberus.ca (Gordon Williams)
Date: Mon, 15 Nov 1999 23:04:19 GMT
Subject: Line Benchmarks Tkinter vs. wxPython
References: <01bf2fa7$02992ca0$8a42c3d1@Cyberus.cyberus.ca> <38306E55.62A6D135@callware.com>
Message-ID: <01bf2fb5$43539380$ed55c3d1@Cyberus.cyberus.ca>

Hi,

See the files in:
http://starship.python.net/pipermail/wxpython-users/1999-November/date.html

with a thread of the same name.

Regards,

Gordon

Ivan Van Laningham <ivanlan at callware.com> wrote in article
<38306E55.62A6D135 at callware.com>...
> Hi All--
> 
> Gordon Williams wrote:
> > 
> > Hello All,
> > 
> 
> [snip]
> 
> > 3)      wxPython beats Tk by a very! large margin even when Tk is
called
> > directly.  Direct drawing onto to screen is roughly 3 times faster than
> > using a bitmap.
> > 
> > Thanks to:
> > Greg Landrum for providing some information on direct calls to tk.
> > Ionel Simionescu for providing the wxPython code which I was able to
> > modify.
> > 
> 
> Could you post the benchmark code?  Or provide us with a link where it
> can be downloaded?
> 
> Thanks,
> Ivan
> ----------------------------------------------
> Ivan Van Laningham
> Callware Technologies, Inc.
> ivanlan at callware.com
> ivanlan at home.com
> http://www.pauahtun.org
> See also: 
> http://www.foretec.com/python/workshops/1998-11/proceedings.html
> Army Signal Corps:  Cu Chi, Class of '70
> Author:  Teach Yourself Python in 24 Hours
> ----------------------------------------------
> 
> 
> 



From framiere at netcom-service.com  Mon Nov 15 10:44:42 1999
From: framiere at netcom-service.com (Florent Ramière)
Date: Mon, 15 Nov 1999 16:44:42 +0100
Subject: PyApache
References: <80p6p4$or5$1@oceanite.cybercable.fr>
Message-ID: <80p9uj$43f$1@oceanite.cybercable.fr>

Hello,

    I have the same problem !
    Does anyone found a solution for this problem, or does pyApache
absolutly do not handle such a functionnality (maybe) ?

Florent.

leon <leonar_n at epita.fr> wrote in message
news:80p6p4$or5$1 at oceanite.cybercable.fr...
> Hi,
>
> I'm using PyApache, and I would like to know if it's possible to make
> 'persistent' modules. In fact I want to do only one "import file.py" like
in
> mod_perl, instead of loading this file each time the script is parsed.
>
> Thanks.
>
>





From stiller at phil.uni-erlangen.de  Fri Nov 26 22:32:33 1999
From: stiller at phil.uni-erlangen.de (Jan Stiller)
Date: Sat, 27 Nov 1999 04:32:33 +0100
Subject: Sending files to a WWW-Server
Message-ID: <81nje2$su9$1@rznews.rrze.uni-erlangen.de>

Hi all

I'm searching for examples how to send files to a web server. I've found
out, that I have to use multipart messages, but don't know exactly how to
construct them.

I'm writing 3-tier applications and really would love to test my backend
code via port 80.

thanks Jan





From gmcm at hypernet.com  Wed Nov 24 08:05:25 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 24 Nov 1999 08:05:25 -0500
Subject: Python-URL! - The weekly guide to Python resources (Nov 23)
Message-ID: <1268673771-3409120@hypernet.com>

It's been a relatively quiet week on Pythondom.
 Those of us in the States, of course, are 
 mentally preparing ourselves for an onslaught
 of fowl, football and the imminent arise of
 Christmas decorations.

We have an announcement form Ben Gertzfield of
 CDDB.py, a program to fetch information about 
 CD audio tracks online (Linux only; GPL'd).
  http://www.deja.com/getdoc.xp?AN=551850359

The main emphasis this week seems to be on 
Python internals, object models and parse
trees, oh my.

One thread has Greg Ewing continuing his 
 exploration of Ruby's internals (and 
 contrasting them with Python). This week
 it's object models:
  http://www.deja.com/getdoc.xp?AN=549499675
  http://www.deja.com/getdoc.xp?AN=551536349

John Skaller talks about the same subject, 
 showing how Viper (his pre-release Python-like
 language) deals with the same issues:
  http://www.deja.com/getdoc.xp?AN=551101329

In an unrelated thread, Greg Ewing provides
 some hints on how to extend Python's grammar:
  http://www.deja.com/getdoc.xp?AN=551856547

Michael Hudson announces release 0.60 of
 his perverse bytecodehacks:
  http://www.deja.com/getdoc.xp?AN=551614919

While John Aycock previews an alpha Python
 decompiler:
  http://www.deja.com/getdoc.xp?AN=550109419

Along the same track, David Jeske announces 
 that work has started on pylint.py, a LINT
 for Python focussing on type-inference:
  http://www.deja.com/getdoc.xp?AN=551464214

=============================================

Everything you want is probably one or two clicks away in 
these pages:

      Python.org's Python Language Website is the center of 
      Pythonia      http://www.python.org

      Much of Python's real work takes place on Special-
      Interest Group  mailing lists http://www.python.org/sigs/

      Python Consortium emerges as an independent nexus of 
      activity   http://www.python.org/consortium

      Python To-Do List anticipates some of Python's future 
      direction  http://www.python.org/cgi-bin/todo.py

      Python Journal is at work on its second issue
          http://www.pythonjournal.com

      Links2Go is a new semi-automated link collection; it's 
      impressive what AI can generate
          http://www.links2go.com/search?search=python

    Archive probing trick of the trade:

http://www.dejanews.com/dnquery.xp?QRY=&DBS=2&ST=PS
&defaultOp=AND&LNG=AL
L&format=threaded&showsort=date&maxhits=100&groups=co
mp.lang.python


Previous - (U)se the (R)esource, (L)uke! - messages are listed 
here:
  http://purl.org/thecliff/python/url.html
or
http://www.dejanews.com/dnquery.xp?QRY=~g%20comp.lang.
python%20Python-URL%21


Suggestions/corrections for next week's posting are always 
welcome.
  http://www.egroups.com/list/python-url-leads/

To receive a new issue of this posting in e-mail each Monday 
morning, ask <claird at neosoft.com> to subscribe.  Mention 
"Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to 
participate in and sponsor the "Python-URL!" project.
-- 

Cameron Laird           
http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX


- Gordon



From bgdarnel at my-deja.com  Tue Nov 23 23:46:02 1999
From: bgdarnel at my-deja.com (Ben Darnell)
Date: 24 Nov 1999 04:46:02 GMT
Subject: PalmPython?
References: <FLooAG.MqL@world.std.com>
Message-ID: <slrn83mrca.cih.bgdarnel@bgdarnel.dhis.org>

In article <FLooAG.MqL at world.std.com>, Will Ware wrote:
>Is PalmPython still around? I think Bob Tillotson (sp?) had been
>working on it. I picked up a Palm Pilot today, and have gotten
>pilot-link working on my Linux box. PalmPython would be a cool
>next thing to get working, if it still exists. Thanks for any
>pointers.

Rob Tillotson's Pyrite (name changed because of Palm's trademark) can be
found at http://pyrite.linuxbox.com/pyrite/

-Ben

-- 
Ben Darnell              bgdarnel at unity.nc su.edu
http://www4.nc su.edu/~bgdarnel/thoughtstream/




From tony at lsl.co.uk  Tue Nov  9 10:33:21 1999
From: tony at lsl.co.uk (Tony J. Ibbs (Tibs))
Date: Tue, 9 Nov 1999 15:33:21 -0000
Subject: [Python] header (fwd)
Message-ID: <209f01bf2ac7$c19b1d50$f0c809c0@lslp7o.lsl.co.uk>

Robert Meegan <rm at murl.com> wrote (in answer to Benoit Ciceron):
>If what you want is a way to filter mailing list messages, may I
>suggest that you look for "python-list" in both the "To:" and "CC:"
>fields.

Unfortunately, demonstrably insufficient (i.e., some messages to the list do
not actually have python-list in either To: or CC: fields). This is a
problem to those (for instance) who are unlucky enough to have to use
Outlook Express (<fx spitting>phthwip-ping</>) which only allows filtering
on To, From, CC and Subject fields. Personally, I just put up with some
messages getting misfiled to my inbox, and dream of one day getting back to
a decent mail handler...

Tibs
--
Tony J Ibbs (Tibs)      http://www.tibsnjoan.demon.co.uk/
Feet first with 5 wheels...
My views! Mine! Mine! (Unless Laser-Scan ask nicely to borrow them.)





From mhammond at skippinet.com.au  Mon Nov 22 19:12:28 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Tue, 23 Nov 1999 11:12:28 +1100
Subject: Getting PythonWin debugger working with an embedded app.
In-Reply-To: <000a01bf3507$6600b9b0$3acbd9c2@peridot.optichrome.com>
Message-ID: <001501bf3559$23860390$0501a8c0@bobcat>

> I would like to suggest to Mark Hammond, if possible, to include
> debug versions of the win32all modules in the distribution, maybe
> as an optional component in the installer.

These are available only to registered users.  See
http://starship.python.net/crew/mhammond

Mark.




From wayne_strauss at mail.medscape.com  Tue Nov  9 13:00:37 1999
From: wayne_strauss at mail.medscape.com (wayne_strauss at mail.medscape.com)
Date: Tue, 9 Nov 1999 13:00:37 -0500
Subject: Newbee WinNT ODBC Question
Message-ID: <85256824.006309BF.00@notessmtp1.medscape.com>


Aaaaaaaargh!
Thanks.




"Gordon McMillan" <gmcm at hypernet.com> on 11/09/99 10:46:47 AM

Please respond to gmcm at hypernet.com

To:   wayne strauss/Medscape at Medscape, python-list at python.org
cc:

Subject:  Re: Newbee WinNT ODBC Question




wayne_strauss writes:

> I've installed Python and Win32all.exe. When I run the following
> (only the first three lines shown):
>
> import dbi, odbc     # ODBC modules
> import time          # standard time module
>
> dbc = odbc.odbc('db/id/pw')
>
> I get:
>
> >>> Traceback (innermost last):
>   File "d:\Program
> Files\Python\Pythonwin\pywin\framework\scriptutils.py", line 237,
> in RunScript
>     exec codeObject in __main__.__dict__
>   File "D:\dev\Python\ODBC\odbc.py", line 1, in ?
>     import dbi, odbc     # ODBC modules
>   File "odbc.py", line 4, in ?
>     dbc = odbc.odbc('db/id/pw')
> TypeError: call of non-function (type module)
>
> If the import worked then why does the actual function call fail?

You didn't by any chance name this script "odbc.py" did you?

import-Narcissus-ly y'rs

- Gordon









From python-list at teleo.net  Mon Nov  8 13:57:11 1999
From: python-list at teleo.net (Patrick Phalen)
Date: Mon, 8 Nov 1999 10:57:11 -0800
Subject: A Modest Proposal -- <wink>
In-Reply-To: <8072o6$m4i$1@newshost.accu.uu.nl>
References: <99110613580407.02970@quadra.teleo.net> <8072o6$m4i$1@newshost.accu.uu.nl>
Message-ID: <99110811064800.01212@quadra.teleo.net>

[Martijn Faassen, on Mon, 08 Nov 1999]:

:: To adopt the <wink> tag some minor changes to XML need to be made. The
:: <wink> tag affects any _previous_ (not containing) markup. I am
:: proposing a system called XWTSNTSDL (XML Wink Tag Schedule Namespace
:: Transformation Semantic Disruption Language) to deal with this. XWTSNTSDL
:: depends on the two other proposals I sent to you previously: XGNBAML and
:: XACL (XML Grand Name But Actually MeaningLess and XML Acronym Complexification
:: Language).

This is bold! I admit I hadn't dared to consider this approach. In one
mighty stroke, it is preservative of the whole Grand Archive of Timbot
Postings (which, as prior art and international intellectual treasure,
certainly deserve any effort necessary).

how-do-you-say-"And-the-tail-shall-wag-the-dog"-in-Dutch?-ly y'rs 
-patrick



From claird at starbase.neosoft.com  Sun Nov  7 14:51:22 1999
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 7 Nov 1999 19:51:22 GMT
Subject: Printing from Python to Windows Printers
References: <fIHU3.641$f66.886@198.235.216.4> <804hb0$4cb6@SGI3651ef0.iddeo.es>
Message-ID: <AFCAF2022974E7E7.1C29CCB3478DD2EB.23272FA059A66374@lp.airnews.net>

In article <804hb0$4cb6 at SGI3651ef0.iddeo.es>, Juan <jlle at redestb.es> wrote:
>You can use the class wxPrinter and wxPrintout instaling previously
>wxPython.
>
>Un saludo.
>Juan Huertas
>
>Robert Parker <rparker at bserv.com> escribi? en el mensaje de noticias
>fIHU3.641$f66.886 at 198.235.216.4...
>> I am trying to find out how I can print to a windows printer from a Python
>> application.  Has anyone used any of the Com modules to gain access to the
>> printer api in windows?  Are there any modules that have been written that
>> would help save time when we want to send font information only to the
>> printer driver not have to write in the escape sequences and print
>directly
>> to the parallel port?
			.
			.
			.
You might want to read
<URL:http://x47.deja.com/=dnc/getdoc.xp?AN=544849327>
which hints at brighter prospects ahead.  The Tclers
are willing at least to consider co-operation with
Pythonists.  Printing looks to me like a domain where
the two communities might work quite effectively to-
gether.
-- 

Cameron Laird           http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX



From pearu at ioc.ee  Tue Nov  9 14:05:59 1999
From: pearu at ioc.ee (Pearu Peterson)
Date: Tue, 09 Nov 1999 21:05:59 +0200
Subject: PyArg_ParseTupleAndKeywords problem
Message-ID: <38287096.689640FD@ioc.ee>

Hi!

I am writing a C/API function that has optional arguments. What I would
like to obtain is a C/API version of the following Python function (just
for example):

def foo(a,b=1):
    return a,b

So, I write the following C code:
<snip>
static PyObject *test_foo(PyObject *capi_self, PyObject *capi_args,
PyObject *capi_keywds) {
    PyObject *a_capi = NULL;
    PyObject *b_capi = NULL;
    static char *capi_kwlist[] = {"b",NULL};
    if
(!PyArg_ParseTupleAndKeywords(capi_args,capi_keywds,"O|O",capi_kwlist,&a_capi,&b_capi))
goto capi_fail;
    return Py_BuildValue("OO",a_capi,b_capi);
capi_fail:
    return NULL
}
<snip>

After building a module I can do in Pyhton:

>>> import test
>>> test.foo(1,4)
(1,4)
>>> test.foo(2)
(2,1)

but that is all!!!!!!
For example,

>>> test.foo(1,b=4)
Traceback (innermost last):
  File "<string>", line 1, in ?
TypeError: keyword parameter b redefined
>>> test.foo(a=1,b=4)
Traceback (innermost last):
  File "<string>", line 1, in ?
SystemError: number of items in format string and keyword list do not
match

etc.
Note that if

>>>def foo(a,b=1): return a,b

then
    foo(1,b=4)
    foo(a=1,b=4)
    foo(b=4,a=1)
are correct.

So, clearly PyArg_ParseTupleAndKeywords command emulates only a subset
of Python possibilities.
Can anyone point me out how to get 'test.foo(1,b=4)' to work?
Should I write my own PyArg_ParseTupleAndKeywords function for that?

Thanks,
    Pearu

--
Pearu Peterson , MSc, Researcher
Department of Mechanics and Applied Mathematics          http://koer.ioc.ee/~pearu/
Institute of Cybernetics at Tallinn Technical University Phone: (+372) 6204168
Akadeemia Rd. 21, 12618 Tallinn ESTONIA                  Fax:   (+372) 6204161
*** the nonvalidity of rigorous causality is necessary
    and not just consistently possible (Heisenberg, 1925) ***






From mhammond at skippinet.com.au  Mon Nov 15 17:25:50 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Mon, 15 Nov 1999 22:25:50 GMT
Subject: win32api... another question:  GetEnvironmentVariable("PATH")
References: <01BF2F61.1D2B8DB0@exchange1.geometria.hu> <W8WX3.67146$Rx2.793309@news1.rdc1.bc.home.com>
Message-ID: <OL%X3.4846$Xo.25269@news-server.bigpond.net.au>

Kevin Cazabon wrote in message ...
>Interesting... it doesn't work on Win98 for me (The exact statement
>below...).

Eeek - you are correct - it doesnt :-(

win32api implements this using "GetEnvironmentVariableW()", and returns a
Unicode string.  This doesnt work on 95/98.  I dont know what I was thinking
:-(

I think I should drop this back to an Ascii version, and return a normal
PyString object.  That shouldnt break any code as Unicode objects are pretty
useless now and need to be "str()"d to be useful.

Damn...

Mark.






From robin at jessikat.demon.co.uk  Mon Nov  8 10:50:22 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Mon, 8 Nov 1999 15:50:22 +0000
Subject: pythonistas or pythoneers?
References: <07111999.2@sanctum.jae>
Message-ID: <E6hsjFA+EvJ4EwiZ@jessikat.demon.co.uk>

In article <07111999.2 at sanctum.jae>, Juergen A. Erhard <jae at ilk.de>
writes
>Hash: SHA1
>
>I'm confused: are we pythoneers or pythonistas?
>
>I tend to go for pythoneers... sounds more pythonic.
>
>Or are these two levels of experience?  If so, what's higher?
>
>Confused,
>
>Bye, J
>
>PS: Of courthe, my thongue wath firmly lodthed in my theek while I
>wrote thith...
>
I prefer pythoneer; it sounds as though we're in charge. A pythonista
sounds like a follower of some revolutionary cult. On the other hand
Zopista sounds right somehow. 
-- 
Robin Becker



From boud at rempt.xs4all.nl  Tue Nov 30 02:12:20 1999
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 30 Nov 1999 07:12:20 GMT
Subject: "Import error"--MySQLmodule broke betw. Python 1.5.1 and 1.5.2
References: <38418F71.A39C213E@dds.nl> <81s70o$1nne$1@hub.org>
Message-ID: <81vtck$g8$4@essle.valdyas.org>

tim muddletin <tim muddletin at news.vex.net> wrote:
> On Sun, 28 Nov 1999 21:24:17 +0100, Martian <martian1 at dds.nl> wrote:
>>The MySQLmodule hasn't changed--indeed, hasn't since Oct. 1998

> Is there any reason you need to use the old module? According to my
> files ( http://www.vex.net/~x/parnassus/apyllo.py?i=82701837 ), there
> seems to be a newer (and quite improved sounding) MySQL module here...

>  http://starship.python.net/crew/adustman/MySQLdb.html

> (note: i don't know anything about it, never having used MySQL.
> I'm just a humble python resource librarian. <-: Though i don't
> seem to have record of the old module in my database! Hmmm...)

The interface is rather different, as is the format of the
returned rows of data. So making the switch can be a lot
of effort.

-- 

Boudewijn Rempt  | http://denden.conlang.org



From paul.m at yale.edu  Tue Nov 16 14:11:15 1999
From: paul.m at yale.edu (Paul M)
Date: Tue, 16 Nov 1999 14:11:15 -0500
Subject: wxPython and DISLIN work nicely together
References: <80s0b4$g5n@mail.psy.uva.nl> <80s60j$gkq@mail.psy.uva.nl>
Message-ID: <80sa7u$9sb$1@news.ycc.yale.edu>

Huh? How did you determine that?  I've used the DISLIN python
extension on FreeBSD (and probably by extension all the *BSD
platforms) without any problems and I would guess it also works on all
the supported UNIX platforms (see the README.UNX file at the DISLIN
site).

--Paul

>
>I have to cut some from "quite a lot of platforms".
>The DISLIN Python extension is only available for Win32 and Linux.
>
>ionel
>
>
>




From S.I.Reynolds at cs.bham.ac.uk  Mon Nov 29 16:06:31 1999
From: S.I.Reynolds at cs.bham.ac.uk (Stuart Reynolds)
Date: Mon, 29 Nov 1999 21:06:31 +0000
Subject: __getslice__
References: <199911292036.VAA14482@axil.hvision.nl>
Message-ID: <3842EAD7.13B9@cs.bham.ac.uk>

Hans Nowak wrote:
> 
...
> Did you try UnaryFnSequence(square, [2,3,4])? You'll get a surprising
> result:
> 
> ((0,0), (1,1), (2,4))
> 
> >     def __repr__(self):
> >         s='( '
> >         for x in range( len(self) ):
> >             s = s+'('+`x`+','+`self._fn(x)`+'), '
> >         return s[:-2] + ' )'
> 
> This is the culprit... You take a list of N elements, and then you
> loop over *a range 0..N-1*... rather than the list! Try replacing the
> 
>   for x in range(len(self)):
> 
> with
> 
>   for x in self._inputs:
> 
> This might do what you intended. ^_^
> 
> Hope this helps,

Nice one.

Cheers,

Stu



From foxes at theriver.com  Fri Nov 12 10:04:10 1999
From: foxes at theriver.com (Andy Fox)
Date: Fri, 12 Nov 1999 08:04:10 -0700
Subject: '==' vs. 'is' behavior
Message-ID: <382C2C6A.64A8F2C2@theriver.com>

Hi,

I thought I understood the difference between the '==' and 'is'
operators, but just found an exception to the rule.  Couldn't find an
explanation in the FAQs.

I understand the following:
>>> a = (1, 2)
>>> b = (1, 2)
>>> a == b #are they equivalent?  yes.
1
>>> a is b #are they the same object? no.
0

Why doesn't it work for a certain integers:
>>> a = 2
>>> b = 2
>>> a == b #are they equivalent?  yes.
1
>>> a is b #are they the same object? yes.  Why?
1

Further experimentation shows that this seems to work for integers
through 99, but that starting with 100 the results of the 'is' test
return 0, as I would expect.

sys.version yields
"Python 1.5.2 (#0, Apr 13 1999, 10:51:12) [MSC 32 bit (Intel)] on win32"

-- 
Andy Fox



From huiray at gate.sinica.edu.tw  Mon Nov 29 12:35:34 1999
From: huiray at gate.sinica.edu.tw (SINICA)
Date: Tue, 30 Nov 1999 01:35:34 +0800
Subject: Access violation while trying to replace stdout with a local disk file
Message-ID: <81udcl$nm3$1@news1.sinica.edu.tw>

Hi there !

I'm trying to redirect the output of Python into a external diskfile
since my program and python interpretor are two different processes
(python15.dll and myprog.exe)
PyObject_Print() will cause a "Access Violation" while trying to
print on the file handle which is created by myprog.exe

any suggestion will be greatly appreciated.

please cc to csyap at pd.jaring.my
thank you in advance.

yap

=====================SAMPLE CODE========================
FILE *output;
PyObject *new_stdout;

output = fopen("adiskfile","w");

new_stdout = PyFile_FromFile(output, "<stdout>", "w", NULL);
PySys_SetObject("stdout", new_stdout);

fprintf(output,"testing");    // <- it works

 v = PyRun_String("print 1+2", Py_single_input, globals, globals);   // <-
it doesn't works







From meowing at banet.net  Tue Nov  2 14:45:04 1999
From: meowing at banet.net (greg andruk)
Date: Tue, 2 Nov 1999 14:45:04 -0500 (EST)
Subject: cStringIO versus StringIO: "readlines"
In-Reply-To: <381F338E.D3D2E748@bioreason.com>
Message-ID: <Pine.LNX.4.10.9911021438190.6763-100000@slip-32-101-160-155.ma.us.prserv.net>

On Tue, 2 Nov 1999, Brian Kelley wrote:

> I noticed that StringIO has a "readlines" member function and cStringIO
> doesn't.  Is anyone planning on adding this in the near future?  I was
> wondering if I should go ahead and do it or if there is against.

You're welcome to try (or borrow parts of) the thingy module at
<URL:http://members.xoom.com/meowing/python/> .  It's got the methods
cStringIO is missing, lets you modify objects created from strings, and
has list methods so you can insert and delete in the middle [though
figuring out what tell() means after that is your problem =)].  Yes, I
actually had a need for such a mongrel, and it's been chugging away with
no trouble (or none that I've noticed anyway!) for a few weeks here.




From tim_one at email.msn.com  Wed Nov 10 00:30:10 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 10 Nov 1999 00:30:10 -0500
Subject: .dll and .exe file version informaton
In-Reply-To: <806tck$5o9$1@nnrp1.deja.com>
Message-ID: <000101bf2b3c$a7d43e40$d32d153f@tim>

[emuller at painewebber.com]
> Can anyone give me an idea how to quikly get the version info on .dll
> and .exe files under windows?

In Explorer, right-click on the file, select Properties from the context
menu, and click the Version tab.

go-ahead-just-*try*-to-write-a-program-to-do-that<wink>-ly y'rs  - tim





From piet at cs.uu.nl  Wed Nov 17 05:56:21 1999
From: piet at cs.uu.nl (piet at cs.uu.nl)
Date: 17 Nov 1999 11:56:21 +0100
Subject: Need help dealing with spaces in NT paths.
In-Reply-To: "T. C. Mits"'s message of "Wed, 17 Nov 1999 01:00:18 -0500"
References: <80pr76$jq5$1@nnrp1.deja.com> <11f733ec.96958819@usw-ex0101-007.remarq.com> <uogcu93dt.fsf@cs.uu.nl> <80sj4q$l7t$1@nnrp1.deja.com> <80tgi8$hbn$1@ssauraaa-i-1.production.compuserve.com>
Message-ID: <uzowdfktm.fsf@cs.uu.nl>

>>>>> "T. C. Mits" <71351.356 at compuserve.com> (TCM) writes:

TCM> A simple example of the difficulty is:
>>>> os.system(r'"dir Program Files"')
TCM> Bad command or file name
TCM> 0
>>>> 


TCM> This is run at the root of C: for simplicity.


TCM> I also get:

TCM> os.popen(r'"dir Program Files"', "r")
TCM> Bad command or file name
TCM> <open file '"dir program files"', mode 'r'  at 795930>

TCM> What does that mean?

It means there is no program with the name: dir program files
You should use:

os.system(r'dir "Program Files"')
os.popen(r'dir "Program Files"', "r")
-- 
Piet van Oostrum <piet at cs.uu.nl>
URL: http://www.cs.uu.nl/~piet [PGP]
Private email: Piet.van.Oostrum at gironet.nl



From billtut at microsoft.com  Mon Nov 15 11:40:12 1999
From: billtut at microsoft.com (Bill Tutt)
Date: Mon, 15 Nov 1999 08:40:12 -0800
Subject: Python ISAPI
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8100DB91555@RED-MSG-50>


> -----Original Message-----
> From: Florent Rami?re [mailto:framiere at netcom-service.com]
> 
> 
> Hello,
> 
>     would it be a good thing to make python ISAPI compliant :) ?

Only in as much as it would resemble mod_perl for Apache.
Using ISAPI with Python in hopes of getting better performance isn't
terribly likely....
(Python's threading model will byte you here.)

Bill




From charliehubbard76 at yahoo.com  Thu Nov  4 12:44:22 1999
From: charliehubbard76 at yahoo.com (Charlie Hubbard)
Date: Thu, 04 Nov 1999 12:44:22 -0500
Subject: Zope comparisons? (ASP, PHP, etc)
Message-ID: <3821C5F6.84543AF2@yahoo.com>

Hello,

I'm looking for some comparison info about zope.  Why zope?
How is it different, the same, or better than ASP, PHP, or
DHTML?  I've looked over some different docs on zope.org,
but I wasn't able to find any info about what need zope fills.
What can you do with zope that you can't do with ASP or PHP?
It seems zope is based off of an object model where PHP and
ASP are simplying server side scripting languages.  Does zope
fill that need?  Is it in competition with this languages?  Would
there be any reason to use Zope with ASP or PHP?

thanks
charlie




From phd at phd.russ.ru  Thu Nov 18 08:43:45 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Thu, 18 Nov 1999 13:43:45 +0000 (GMT)
Subject: My Python software: URL checker; netscape history export/import
Message-ID: <Pine.LNX.4.21.9911181335001.27748-100000@fep132.fep.ru>

Hello!

   I am announcing some updates on my software page:
http://sun.med.ru/~phd/Software/Python/
   and mirrors:
http://members.xoom.com/phd2/Software/Python/
http://www.fortunecity.com/skyscraper/unix/797/Software/Python/

   There is new version of Internet robot (URL checker) in the "Bookmarks
database and Internet robot". The robot now forking off child process,
which does real job; parent controls the child and kills it if 15 timeout.

   There is new tarball in "misc" - netscape_history.tgz. There is
Navigator's history dmper - a slightly modified version of Hannu Krosing's
original program. And, of course, the reverse program to import text file
back into bsddb (Berkeley DB 1.85) format.

Oleg.
---- 
    Oleg Broytmann  National Research Surgery Centre  http://sun.med.ru/~phd/
           Programmers don't die, they just GOSUB without RETURN.





From kcazabon at home.com  Mon Nov 15 11:03:02 1999
From: kcazabon at home.com (Kevin Cazabon)
Date: Mon, 15 Nov 1999 16:03:02 GMT
Subject: win32api... another question:  GetEnvironmentVariable("PATH")
References: <01BF2F61.1D2B8DB0@exchange1.geometria.hu>
Message-ID: <W8WX3.67146$Rx2.793309@news1.rdc1.bc.home.com>

Interesting... it doesn't work on Win98 for me (The exact statement
below...).

Anyone have suggestions for Win95/98?


Kiss, Arpad <AKiss at GEOMETRIA.hu> wrote in message
news:01BF2F61.1D2B8DB0 at exchange1.geometria.hu...
> Hi,
>
> Kevin Cazabon <kcazabon at home.com> wrote in message
> news:<fcMX3.65012$Rx2.773564 at news1.rdc1.bc.home.com>...
> > I can't seem to get the win32api.GetEnvironmentVariable() function to
> work
> > properly... am I not using it correctly by the following?
> >
> > path = win32api.GetEnvironmentVariable("PATH")
>
> It works for me on a Windows NT Workstation (sp5).
>
> >
> > or
> >
> > path = win32api.GetEnvironmentVariable("%%PATH%%")
>
> It doesn't work....
>
>
>
>





From ivnowa at hvision.nl  Tue Nov 30 17:44:59 1999
From: ivnowa at hvision.nl (Hans Nowak)
Date: Tue, 30 Nov 1999 23:44:59 +0100
Subject: performance of tuple-less Python?
In-Reply-To: <slrn848jos.nnk.quinn@amatho.ugcs.caltech.edu>
Message-ID: <199911302244.XAA22802@axil.hvision.nl>

On 30 Nov 99, Quinn Dunkan wrote:

> And what was that party trick of changing 1 to equal 2 in forth?  Maybe
> that's more at the language level though, but I don't know much about
> forth (one of these days...).

Easy... let's change the value of 4:

: 4 5 ;
4 4 + .

will print 10. ^_^

Python is part OO, part procedural, part functional... I haven't seen 
any features from stack-based languages yet, though. Maybe an idea 
for Guido? <wink>

--Hans Nowak (zephyrfalcon at hvision.nl)
Homepage: http://fly.to/zephyrfalcon
You call me a masterless man. You are wrong. I am my own master.



From donn at oz.net  Fri Nov 26 13:57:01 1999
From: donn at oz.net (Donn Cave)
Date: 26 Nov 1999 18:57:01 GMT
Subject: fcntl/ioctl and O_NONBLOCK
References: <slrn83t5mv.bhk.hildeb@stahlw06.stahl.bau.tu-bs.de>
Message-ID: <81ml5t$eer$0@216.39.151.169>

Quoth hildeb at www.stahl.bau.tu-bs.de (Ralf Hildebrandt):
... [stdout is Popen3.fromchild]

| rv = fcntl.ioctl(stdout.fileno(), FCNTL.O_NONBLOCK, 1)
| # rv = fcntl.ioctl(stdout.fileno(), FCNTL.O_NDELAY, 1)
| ... snap ...
|
| But no matter what I do - I keep getting
| IOError: [Errno 22] Invalid argument
|
| What I want to do: I want to use select() to see if there's data to be
| read from stdin or stdout and read it ALL (thus I need stdin/stdout to
| be non blocking!)

I don't know about the ioctl, but you may not need it after all if
you're already going to use select.  Select will tell you if the
I/O would block, if so then don't read.

(And you can't read from both ends, "stdin" (i.e., Popen3.tochild)
is opened for write and attempts to read from it will probably be
frustrating.)

	Donn Cave, donn at oz.net



From news at dorb.com  Sun Nov 28 14:40:44 1999
From: news at dorb.com (Darrell)
Date: Sun, 28 Nov 1999 14:40:44 -0500
Subject: is there a safe/easy way to search and replace text w/in an html page?
References: <3840593B.2D1A2E30@earthlink.net>
Message-ID: <bxf04.15814$ss3.618968@typhoon.nyroc.rr.com>

The attached code isn't very well tested.
And I'm sure as always that someone else will point out that this was
possible
from the lib :)

--
--Darrell

"Levi Cook" <levicook at earthlink.net> wrote in message
news:3840593B.2D1A2E30 at earthlink.net...
> Hello,
>
> I'm new to python and I've been tasked w/ writing a script that fetches
> a web-page and highlights the term that the user searched for. Passing
> the URL and search term to my script was a cinch and fetching the page
> was also quite easy.
>
> The crux of my problem lies with establishing a strategy for searching
> out the users term and replacing it with css tags surrounding the
> original term.
> At first glance the re module seemed sufficient (and maybe still is ),
> but it doesn't really have any knowledge of HTML semantics. My next
> venture was to look at the htmllib module. This appears to parse the
> html fine, but i'm still at a bit of a loss on what this actually does
> for me, particularly how do I get my parser to output my page.
>
> Does anyone know where I can find code that does anything remotely like
> this?
> Does anyone know where I can find good reference on the htmllib module,
> the python library reference is pretty bleak in this area.
>
> Thanks in advance for you time and help.
> Levi Cook
> levicook at earthlink.net



begin 666 url.py
M:6UP;W)T('5R;&QI8BP@<F4L(&AT='!L:6(L(&)I<V5C= T*#0I?7W-E<4YU
M;3TP#0ID968 at 7V9I;F1A;&Q%>'0H<&%T=&5R;BP@<V]U<F-E+"!S=&%R=#TP
M+"!S=&]P/3 I. at T*(" @(" @(" B(B)2971U<FX at 82!L:7-T(&]F(&%L;"!N
M;VXM;W9E<FQA<'!I;F<@;6%T8VAE<R!I;B!T:&4@<W1R:6YG+ at T*#0H@(" @
M(" @($EF(&]N92!O<B!M;W)E(&=R;W5P<R!A<F4@<')E<V5N="!I;B!T:&4@
M<&%T=&5R;BP@<F5T=7)N(&$-"B @(" @(" @;&ES="!O9B!G<F]U<',[('1H
M:7,@=VEL;"!B92!A(&QI<W0@;V8@='5P;&5S(&EF('1H92!P871T97)N#0H@
M(" @(" @(&AA<R!M;W)E('1H86X@;VYE(&=R;W5P+ at T*#0H@(" @(" @($5M
M<'1Y(&UA=&-H97, at 87)E(&EN8VQU9&5D(&EN('1H92!R97-U;'0N#0H-"B @
M(" @(" @4F5T=7)N<R!A(&QI<W0@;V8 at 6RAS=&%R="P@<V5Q3G5M8F5R+"!E
M;F0I+"XN+ET-"@T*(" @(" @(" ^/CX@<S$])V%X8GAC>&%X8GAC>"<-"B @
M(" @(" @/CX^(&9I;F1A;&Q%>'0H)V$N*C]B+BH_8R<L<S$I#0H@(" @(" @
M(%LH,"P@,"P at -2DL("@V+" Q+" Q,2E=#0H@(" @(" @("(B(@T*(" @(" @
M("!G;&]B86P at 7U]S97%.=6T-"B @(" @(" @<&]S(#T@<W1A<G0-"B @(" @
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M(" @(" @(" @(" @(" @(" @(&=R(#T at 6UT-"B @(" @(" @(" @(" @(" @
M(" @(" @(&9O<B H82P at 8BD@:6X@<F5S=#H-"B @(" @(" @(" @(" @(" @
M(" @(" @(" @(" @(" @9W(N87!P96YD*"AA+&(I*0T*(" @(" @(" @(" @
M(" @(" @(" @(" @9W(@/2!T=7!L92AG<BD-"B @(" @(" @(" @(" @("!A
M<'!E;F0H9W(I#0H@(" @(" @(" @(" @(" @7U]S97%.=6T]7U]S97%.=6TK
M,0T*(" @(" @(" @(" @(" @('!O<R ](&UA>"AJ+"!P;W,K,2D-"B @(" @
M(" @<F5T=7)N(')E<W5L=',-"@T*#0ID968 at 9V5T06QL5&%G<RAD871A*3H-
M"B @("!R971U<FX at 7V9I;F1A;&Q%>'0H<F4N8V]M<&EL92 at G/%M>/ETJ/SXG
M*2P at 9&%T82D-"@T*#0H-"G-T87)T5&%G(#T@)SQF;VYT(&-L87-S(#T@:&EG
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M<V5R=" \8F%S93X@<V\@=&AA="!A;&P at 96UB961E9"!54DPG<R!R97-O;'9E
M(&-O<G)E8W1L>0T*=&5R;7-T<FEN9R ]("<H/VUI*5Y<<RH\:&5A9#Y<<RHG
M#0IT97)M<&%T(#T@<F4N8V]M<&EL92AT97)M<W1R:6YG*0T*9&%T82 ]('1E
M<FUP870N<W5B*"<\:&5A9#Y<;CQB87-E(&AR968](B5S(CY<;B5S)UP-"B @
M(" E*&%D9')E<W,L8W-S5&%G*2P at 9&%T82D-"@T*#0H@(" @(R!&:6YD(&%L
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M*"<H8V]M*2<I#0ID871A,3UP870N<W5B*')E<"P at 9&%T82D-"@T*;W!E;B at G
@9&%T82YH=&UL)RPG=R<I+G=R:71E*&1A=&$Q*0T*#0H`
`
end




From thomas at xs4all.nl  Fri Nov  5 09:59:50 1999
From: thomas at xs4all.nl (Thomas Wouters)
Date: Fri, 5 Nov 1999 15:59:50 +0100
Subject: return boolean from functions
In-Reply-To: <7vuoqt$m79$1@nnrp1.deja.com>; from prestonlanders@my-deja.com on Fri, Nov 05, 1999 at 02:16:03PM +0000
References: <7vrq9b$hmr$1@nnrp1.deja.com> <7vsvef$elo$1@nnrp1.deja.com> <7vuoqt$m79$1@nnrp1.deja.com>
Message-ID: <19991105155950.S14322@xs4all.nl>

On Fri, Nov 05, 1999 at 02:16:03PM +0000, Preston Landers wrote:

> and false for fail.  That's fine if that works for you, but FYI the
> convention in Unix[1] is generally to return false for success, and true
> for an error condition.  This way the exact condition can be contained

That depends on what part of UNIX you mean :-) In UNIX shell programming,
where 'true/false' is based on the exit codes of programs, '0' means
success, and something non-zero means failure. But it just means '0' is
true, not that 'false means success'. The UNIX tool 'true' returns '0':

centurion:~ > /bin/true
centurion:~ > echo $?
0

Of course, if you call 'exit(0)' in your program, you are sort-of 'returning'
a false value to signify success -- but it's only false in C evaluations,
not in shell evalutations. Look at it like not 'returning' 0, but 'exiting
with 0 errors'. Or something.

A lot of the UNIX library calls return '-1' for failure, especially when
both '0' and '1' can be valid returns. Some return a code to indicate which
error occured, where 0 might mean 'no error'. Some always return an error,
and some never return :)

conventions-conschmentions'ly y'rs,
-- 
Thomas Wouters <thomas at xs4all.net>

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



From skaller at maxtal.com.au  Tue Nov  2 08:54:07 1999
From: skaller at maxtal.com.au (skaller)
Date: Wed, 03 Nov 1999 00:54:07 +1100
Subject: Assorted Questions regarding the future of Python...
References: <6D8A17398E28D3119F860090274DD7DB49833A@pces.cadlab.it>
Message-ID: <381EECFF.2C04A6EB@maxtal.com.au>

Alessandro Bottoni wrote:

> So, I wonder:
> 1) How about a Python compiler? Does anybody is working on such a project? I
> saw a lot of messages regarding this topic in the past, but which is the
> current status?

	I'm developing a full scale Python program compiler.
The current status is: the compiler works by first 'importing'
modules dynamically, then compiling the resulting dictionaries.
The 'importing' effectively requires a clone of Python, this part
of the compiler is called Viperi [Viper Interpreter]. It will be
available as a standalone tool, and the run-time it uses will
be part of the Ocaml back-end of the compiler.

	The interpreter now executes most Python correctly,
there are a few "not implemented" exceptions, and of course a few
bugs. Performance of the interpreter varies, but is currently
3-10 times slower than CPython (although some operations are actually
much faster than CPython, such as big numbers and loops, and some
are much slower ... ).

	I'm half way through implementing 'binding': freezing
names, and converting them to integers used to access arrays,
instead of dictionaries: CPython does this in functions, it is called
'fast loading'. Viperi 'fast loads' ALL names: whether they're 
function locals, module globals, or class and object attributes.

	[This feature speeds up pystone by about 15-25%]

	Here's what Viper will do: bind all names, merge all modules 
into a single 'address space' and replace A.B access with an access to
A_B. Inline function calls, removing substantial overhead. All these
features will be available in the interpreter, by using special
functions (such as 'bind_function(f)' which creates a bound version 
of a function).

	Now, we do type inference and ad hoc special case
pattern matching, seeking opportunites for optimisation.
Some of these features will be available in the interpreter.

	Finally, the resulting annotated syntax tree
is compiled into the target language: I am guessing
there will be two back ends: an Ocaml (native) back end,
for clients wanting a high power language to extend Python with,
and C, for clients desiring compatibility with CPython extensions.

	I hope to make an announcement soon (as soon as I have
it working well enough to execute 'Interscript'), seeking funding
from the Python and Ocaml communities and prospective clients.
If enough funding is forthcoming, the project will be 'Free for any
use',
otherwise it will have to be shareware or commercial.

> 2) How about a Python Virtual Machine, "a la Java", as a possible
> alternative to a real compiler? Is there anything like that? Is anybody
> working on it?

	The Python VM is an implementation dependent feature already
used by the Python bytecode system to execute bytecode.

-- 
John Skaller, mailto:skaller at maxtal.com.au
1/10 Toxteth Rd Glebe NSW 2037 Australia
homepage: http://www.maxtal.com.au/~skaller
downloads: http://www.triode.net.au/~skaller



From pinard at iro.umontreal.ca  Mon Nov 29 12:45:36 1999
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 29 Nov 1999 12:45:36 -0500
Subject: performance of tuple-less Python?
In-Reply-To: "Greg Wilson"'s message of "Mon, 29 Nov 1999 11:29:30 -0500"
References: <HSx04.8911$cz3.321455@nnrp1.uunet.ca>
Message-ID: <oqso1pdwe7.fsf@titan.progiciels-bpi.ca>

"Greg Wilson" <gvwilson at nevex.com> writes:

> [...] the use of tuples (which I believe are conceptually redundant) [...]

Tuples may be used to index within dictionaries, and I find this feature
very useful.  Lists may not.  I do not know the implementation details,
but would be tempted to have the prejudice (:-) that using lists to index
would imply some severe performance penalty.

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





From mlh at vier.idi.ntnu.no  Mon Nov  8 09:58:31 1999
From: mlh at vier.idi.ntnu.no (Magnus L. Hetland)
Date: 08 Nov 1999 15:58:31 +0100
Subject: pythonistas or pythoneers?
References: <07111999.2@sanctum.jae> <382633B1.50993D93@callware.com>
Message-ID: <y0jn1spni7c.fsf@vier.idi.ntnu.no>

Ivan Van Laningham <ivanlan at callware.com> writes:

> Hi All--
> 
> "Juergen A. Erhard" wrote:
> > 
> > -----BEGIN PGP SIGNED MESSAGE-----
> > Hash: SHA1
> > 
> > I'm confused: are we pythoneers or pythonistas?
> > 
> > I tend to go for pythoneers... sounds more pythonic.
> > 
> > Or are these two levels of experience?  If so, what's higher?
> > 
> 
> Be confused no longer; as we are all guerilla programmers, we are of
> course Pythonistas!

No way - we are all Pythoneers at the frontiers for computer
science... ;)

> 
> <come-the-revolution>-ly y'rs,
> Ivan

--

  Magnus          Echelon jamming noise:
  Lie             FBI CIA NSA Handgun Assault Bomb Drug Terrorism
  Hetland         Special Forces Delta Force AK47 Hillary Clinton 



From fdrake at acm.org  Thu Nov 11 12:32:22 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Thu, 11 Nov 1999 12:32:22 -0500 (EST)
Subject: .html -> .ps ?
In-Reply-To: <19991111122320.A1862@better.net>
References: <19991111080506.A1007@better.net>
	<14378.54900.714487.334690@weyr.cnri.reston.va.us>
	<19991111120757.C1783@better.net>
	<14378.63638.48886.564135@weyr.cnri.reston.va.us>
	<19991111122320.A1862@better.net>
Message-ID: <14378.64934.843002.424264@weyr.cnri.reston.va.us>

William Park writes:
 > I think I can use '-noraise' to prevent the window opening... Yes, just
 > tried it.

  Cool; I thought that only stopped it from raising the window.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From ionel at psy.uva.nl  Mon Nov  1 11:05:50 1999
From: ionel at psy.uva.nl (Ionel Simionescu)
Date: Mon, 1 Nov 1999 17:05:50 +0100
Subject: Just a note
References: <381D5E84.98931682@bibsyst.no> <381D78C3.9BBECCB0@bibsyst.no>
Message-ID: <7vkdoa$q4r@mail.psy.uva.nl>

|
| I will store HUGE amounts of data, so I think Gadfly would be "too small"
| as DBMS, since it stores all it`s data in memory ( as I recall ) .

Just a note.

You can still store huge amounts of data with Gadfly, if you delegate the
storage itself to disk files and keep in Gafly just the filenames.

This scheme, however, would impose on the data access. Namely, you will not
be able to easily apply operators on the full database, but only on the
items you can keep in memory.
And you will need to manage what's kept in memory all by yourself.

However, if you do not need access over the full data domain at the same
time, then you can load data in Gadfly as required and still have a smooth
workflow.

Because this is the usage pattern most of the times, you may still give
Gadfly a thought. Its advantage is better integration with your code
(because Gadfly will do many type conversion for you autmatically) and the
opportunity to hack in Gadfly your specialized operators.

I hope this helps,
ionel









From mehta at mama.indstate.edu  Thu Nov 11 11:56:41 1999
From: mehta at mama.indstate.edu (mehta at mama.indstate.edu)
Date: Thu, 11 Nov 1999 16:56:41 GMT
Subject: Using GNU C complier as the "default" windows complier
References: <4D0A23B3F74DD111ACCD00805F31D8100DB9151B@RED-MSG-50> <80cial$gfn@gap.cco.caltech.edu>
Message-ID: <80esg7$1k9$1@nnrp1.deja.com>

Hello,

are there three simple things available from this gnu c compile
mentioned here form mumit.

1.  compile c programs to binaries.
2.  make (use it as usual)
3.  can i make dll for winnt as I intend to try out one.
4.  could one guide with simple examples.



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



From jon at rdt.monash.edu.au  Mon Nov 29 22:46:34 1999
From: jon at rdt.monash.edu.au (Jonathan Giddy)
Date: 30 Nov 1999 03:46:34 GMT
Subject: Pthreads
References: <81v2an$n6o$1@bgtnsc01.worldnet.att.net>
Message-ID: <81vhaq$cri$1@towncrier.cc.monash.edu.au>

bkenniso at worldnet.att.net (Brian Kennison) writes:

>I'm having trouble compiling Python 1.5.2. with the FSU pthread
>library. I working on MachTen and get the following error:

>gcc -g -O2 -I./../Include -I.. -DHAVE_CONFIG_H   -c thread.c -o thread.o
>In file included from thread.c:150:
>thread_pthread.h:113: parse error before `pthread_cond_t'
>thread_pthread.h:113: warning: no semicolon at end of struct or union
>thread_pthread.h:114: warning: data definition has no type or storage class
>thread_pthread.h:115: warning: data definition has no type or storage class
>thread_pthread.h: In function `PyThread_start_new_thread':       

I'm not sure this is the same problem, but I've seen similar messages under
different circumstances (freezing with extension modules on Solaris).

The problem in that case was that the file Include/thread.h in the Python 
source clashed with the system thread.h file.  Try deleting Include/thread.h 
and recompiling.  (In 1.5.2, thread.h is unused by Python, although it may 
be used by some extensions).

Jon.



From psoares at consiste.pt  Wed Nov 10 06:55:38 1999
From: psoares at consiste.pt (Paulo Soares)
Date: Wed, 10 Nov 1999 11:55:38 -0000
Subject: Syntax colouring in WinEdit?
Message-ID: <51998F46F8B7D011BF7C0060B03502AD2F353C@orion.consiste.pt>

Does anyone has something to do python syntax colouring in WinEdit?

Best Regards,
Paulo Soares



From sbarron at twilight.  Fri Nov  5 23:14:20 1999
From: sbarron at twilight. (Scott Barron)
Date: Sat, 06 Nov 1999 04:14:20 GMT
Subject: Writing to files from C functions
Message-ID: <wWNU3.3454$Gj3.180354@typ11a.deja.bcandid.com>

Hi folks,

I have a C library with a couple of functions along these lines:

int foo (int d, int v)
{
	fd = fopen ("logfile.log", "a");
	fprintf ("called foo on %d with arg %d\n", d, v);
	fflush (fd);
	fclose (fd);
	return 0;
}

I have used SWIG to generate wrappers for this library and all is mostly
well.  However, I'd like to 'tail -f' the log file while im in the Python
interactive interpreter but the data isn't written to the file until I
exit the Python interpreter.  Is Python not actually flushing my file as
I ask or is my C code incorrect?

Thanks,
Scott




From petrilli at digicool.com  Mon Nov 15 10:40:14 1999
From: petrilli at digicool.com (Christopher Petrilli)
Date: Mon, 15 Nov 1999 10:40:14 -0500
Subject: Zope comparisons? (ASP, PHP, etc)
References: <3821C5F6.84543AF2@yahoo.com><06lU3.259$Ae3.186544@news-dal.corridex.com>
Message-ID: <B455938E.5FE6%petrilli@digicool.com>

in article 06lU3.259$Ae3.186544 at news-dal.corridex.com, Evan Simpson at
evan at tokenexchange.com wrote on 11/4/99 2:26 PM:

> 3. ZODB: A versioning, transaction-based object store with automatic object
> persistence and conflict handling.  ZODB stores an object hierarchy as a
> single file (2GB limit being worked on) and appends new versions of objects
> when their request transaction is comitted.  This allows extensive undo, and
> requires occasional packing to eliminate old, unwanted versions.  Most HTML
> is generated by standard Zope objects using...

Just want to clarify that this is a Linux restriction, not a Zope
restriction.  We use a 64-bit OID, so if the underlying OS supports larger
files, then we should continue to grow.

Chris
-- 
| Christopher Petrilli        Python Powered        Digital Creations, Inc.
| petrilli at digicool.com                             http://www.digicool.com




From alexisjuh at usa.net  Wed Nov 10 08:35:03 1999
From: alexisjuh at usa.net (Alexis)
Date: 10 Nov 99 08:35:03 MET
Subject: Alphanumerics
Message-ID: <19991110073503.13073.qmail@nwcst021.netaddress.usa.net>

I'm making some software development for a python program created in the US,
well, it's pretty much just translating and adding some language specific
modules for Norway, Sweden and Denmark...Problem is the special chars we have
here, like ??? for norwegian..
When I use a string everything works fine, but when I import fields from
databases and they contain these chars I'll get escape seq instead...
The chr(num) function doesn't work for these chars for some reason, and now
I'm wondering if there's another way to get it right...
Thanx a lot
Best regards
Ulf (Alexis)

____________________________________________________________________
Get free email and a permanent address at http://www.netaddress.com/?N=1



From pinard at iro.umontreal.ca  Mon Nov  1 20:29:48 1999
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 01 Nov 1999 20:29:48 -0500
Subject: `re' difficulty?
In-Reply-To: "Tim Peters"'s message of "Tue, 19 Oct 1999 21:46:59 -0400"
References: <000601bf1a9c$fff91ec0$252d153f@tim>
Message-ID: <oq3duphe9f.fsf@titan.progiciels-bpi.ca>

"Tim Peters" <tim_one at email.msn.com> ?crit:

> I've been toying for two decades with writing a lengthy but edifying
> treatise revealing the most surprising floating-point numbers I have
> been privileged to know [...]

A was amused, once younger, when someone told me that there are real numbers
that nobody will never speak about.  The reason being simply that there
are more real numbers than there are possible sentences.  (Do not raise
another form of Russell's paradox from within the previous sentence :-).

> [...] but have thus far held back for fear of starving to death.

Have you ever glanced though the (now classic) `On games and numbers' by John
Conway? (the guy behind the `life' game, as you surely know.)  I think the
book sold pretty well.  Who could resist a theory allowing for an infinite
number of distinct zeroes, fuzzy numbers and even wet numbers? :-)

Your turn, Tim!!

P.S. - A detail also much amused me: the book has only two encompassing
chapters, numbered 0 and 1.  Yet, a friend told me that this idea has been
used many times.

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





From b2blink at hotmail.com  Thu Nov 18 11:23:00 1999
From: b2blink at hotmail.com (Ulf Engstrøm)
Date: Thu, 18 Nov 1999 11:23:00 CET
Subject: Python in Windows CE?
Message-ID: <19991118102301.910.qmail@hotmail.com>

>You will probably want to check out the Python CE
>mailing list:
>
>    http://www.egroups.com/group/python-ce/
>
>Python works great on the CE platform. The chief
>shortcoming is the lack of a friendly GUI, but
>this issue is being addressed.  Check the mailing
>list for further details.
>
>-Jeff Bauer

Thanx a lot for this post, this group will come in handy without a doubt. 
After reading some there, looking for an earlier addressed question from me, 
I still wonder how compatible win32-apps are with the CE-version of Python?
Like Mark Hammond write in his post (#4):

While both CE and Python attempt to be portable, it appears that the current 
situation is not at all portable.  It appears there is no way we can write a 
win32 Python program, and have it work on both CE, and Win95/NT.

At the most basic level, simple win32 wrappers are in differently named 
modules.  For example, the CE port has win32reg for the registry functions, 
but in the existing win32 ports, they exist in win32api.  This is not to 
suggest my scheme is anywhere near ideal, but IMO a common strategy is 
needed.
-cut-

Has any solutions to this been found? And are there any other compability 
problems?

Best regards
Ulf Engstr?m

http://www.b2b-link.com
ulf.engstrom at b2b-link.com

______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com



From mlh at vier.idi.ntnu.no  Tue Nov  9 09:12:40 1999
From: mlh at vier.idi.ntnu.no (Magnus L. Hetland)
Date: 09 Nov 1999 15:12:40 +0100
Subject: Stackless python progress?
References: <38274626.5D8C1EC2@ptway.com> <FKwIAv.IB1@world.std.com>
Message-ID: <y0jg0yf21pj.fsf@vier.idi.ntnu.no>

wware-nospam at world.std.com (Will Ware) writes:

> Brian Haskin (haskin at ptway.com) wrote:
> > also has there been any work on combining
> > 'microthreading' and the stackless stuff?
> 
> I was the one doing the microthread stuff. I've been too busy to
> play with it recently. Implementing microthreads in terms of
> Stackless Python is the Right Thing To Do(tm) from an engineering
> standpoint, but I haven't had time to devote to it lately.

How about freezing interpretations? This would be very interesting for
mobile agent systems... (I have already suggested this for JPython).
You could stop a thread of execution, pickle it, and send it to
another location, and then resume it...

--

  Magnus          Echelon jamming noise:
  Lie             FBI CIA NSA Handgun Assault Bomb Drug Terrorism
  Hetland         Special Forces Delta Force AK47 Hillary Clinton 



From emuller at painewebber.com  Mon Nov  8 11:22:50 1999
From: emuller at painewebber.com (emuller at painewebber.com)
Date: Mon, 08 Nov 1999 16:22:50 GMT
Subject: .dll and .exe file version informaton
Message-ID: <806tck$5o9$1@nnrp1.deja.com>

Can anyone give me an idea how to quikly get the version info on .dll
and .exe files under windows?


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



From tismer at appliedbiometrics.com  Tue Nov  9 18:57:51 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Wed, 10 Nov 1999 00:57:51 +0100
Subject: Stackless python progress?
References: <38274626.5D8C1EC2@ptway.com>
Message-ID: <3828B4FF.89B6EA52@appliedbiometrics.com>


Brian Haskin wrote:
> 
> Just wondering if the work on stackless python is still progressing and
> if  it looks like it will be integrated into the regular distribution in
> the foreseeable future? also has there been any work on combining
> 'microthreading' and the stackless stuff?

It is progressing. At the moment I don't publish, but prepare a paper
for IPC8 which has been accepted. Version 0.5 of SP, together with
a quite tough continuation extension module will be published in a
couple of days when the paper is ready.

On microthreads: I've had no more time than a first reading,
and just looked into the latest version right now.
Will, I see you have worked on it, and I appreciate the notes file.
I have been quite away for a while...
Not sure if I have time to discuss this now.
At least so much: that f_callguard was a contribution to make
continuations safe, without building anything into the core
but that one pointer. continuationmodule is a true extension.
If we can isolate the microthread objects into another extension
module and provide a mechanism to get along with this pointer,
would that suffice?
I think to change the semantics of this pointer to be a chain of
hooks, yet not sure about the right protocol.
My current check for something special to happen seems to be cheapest
possible (is that pointer set then call it). But it would not
do any harm if there were one indirection more.

And I see some problems: How do uthreads behave when a call to
a functions happens to create a new interpreter incarnation?
Well, we should discuss this via private email.

will think more of that after the paper is done :-) ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From fredrik at pythonware.com  Sat Nov 20 06:20:43 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 20 Nov 1999 12:20:43 +0100
Subject: import question
References: <O4db2GuM$GA.187@cpmsnbbsa02>
Message-ID: <00c201bf3349$49cc53e0$f29b12c2@secret.pythonware.com>

Bill Wilkinson <yopen at my-Deja.com> wrote:
> I have searched the documentation but have not found anything that helps me
> with this.
> 
> I want to be able to import a module from a  variable instead of a file.  I
> need to do this because all my modules are going to be in a database.  So I
> want have something like this:
> 
> s = "lots of classes and functions loaded from a database"
> import s
> s.run_function()

check out the imputil.py module in greg stein's "small"
distribution kit [1] (also in gordon's installer kit [2]).  it
allows you to create and install custom importers using
only a few lines of Python code.

</F>

1) http://www.lyra.org/greg/small/
2) http://starship.python.net/crew/gmcm/install.html




From rkaiser at wwa.com  Fri Nov 26 21:45:01 1999
From: rkaiser at wwa.com (Ralph Kaiser)
Date: Fri, 26 Nov 1999 20:45:01 -0600
Subject: Problems Finding and installing Tkinter
References: <383ED44D.3E31C959@dimensional.com>
Message-ID: <SyH%3.4511$tI2.92357@ord-read.news.verio.net>

Jim Cochrane <jtc at dimensional.com> wrote in message
news:383ED44D.3E31C959 at dimensional.com...
> I've been trying, with no success, to install the Tkinter package for
> developing GUIs in Python.  First I tried downloading the latest version
> I could find on the net, as well as the latest version of Python, to
> make sure everything was in synch.  I then ran into problems with
> libraries needed for Tkinter.  After downloading and installing several,
> including tkinter, I still had problems and gave up.  (All of these were
> .rpm files.  The last error I remember getting was for an unresovled
> symbol:  __bzero.)
>
> Next I tried uninstalling everything I had installed, and then
> installing the version of Tkinter in my SuSE distribution (and
> reinstalling Python).  (I'm using SuSE 6.1.)  I made sure all the needed
> packages were installed, such as tk and tcl.  Again, no luck.  When
> I try to execute a simple example:
>
> #!/usr/bin/env python
>
> from Tkinter import *
> widget = Label (None, text = 'Hello GUI word!')
> widget.pack()
> widget.mainloop()
>
> I get the following error message:
>
> Traceback (innermost last):
>   File "gui1.py", line 3, in ?
>     from Tkinter import *
>   File "/usr/lib/python1.5/lib-tk/Tkinter.py", line 5, in ?
>     import _tkinter # If this fails your Python is not configured for Tk
>
> ImportError: /usr/X11R6/lib/libtk8.0.so: undefined symbol: panic
>
> So it's not finding a symbol named "panic".  I have no idea what library
> this should be in.
>
> Any help, clues, or pointers would be appreciated.  Also, I think
> another way to do this might be to download and install the latest
> version of everything needed for Tkinter - as I tried to do before - but
> obtaining everything from the same place as one consistent package.
> Does anyone know if this is possible and, if so, where I should look for
> this?


Basically, the _tkinter module in Python is not enabled by default.
Find ../Python-1.5.2/Modules/Setup. In the section marked "The _tkinter
module." you'll find instructuions to enable _tkinter. You might be able to
just make and make install the Modules directory, but I usually do the
entire Python distribution. Good luck.

Ralph






From laurent.duperval at cgi.ca  Tue Nov  2 09:23:31 1999
From: laurent.duperval at cgi.ca (laurent.duperval at cgi.ca)
Date: Tue, 2 Nov 1999 09:23:31 -0500 (EST)
Subject: embedding stubbed tcl
References: <bLd7vSA1euH4EwDS@jessikat.demon.co.uk>
Message-ID: <8ZCT3.18814$1v5.202282@newscontent-01.sprint.ca>

On  2 Nov, Robin Becker wrote:
> Python uses tkinter which embeds tcl and tk. At present python 1.5.2
> comes with a dll that explicitly links to tcl80.dll. having just moved
> to the latest tcl8.2.1 I have to recompile this dll and with a very few
> minor changes all works ok. What happens when 8.3 appears? I guess I
> have to do this again :(
> 

Maybe not.

> What is the simplest way to avoid this pain using stubs? I mean what are
> the guaranteed entry points in the dlls. I need the interp pointer
> before I can start off the stubs interface and it seems that the first
> thing done by tkinter does is Tcl_CreateInterp followed by
> TclpInitLibraryPath(baseName). The latter appears not to be in the stubs
> library. Therefore it seems as though tkinter is not stubs friendly. I
> guess that the embedding structure for tkinter is a bit out of sync with
> the correct way of doing things.

Stubs were not included in the core until 8.1.x and 8.2.x. So your situation
is normal. But if you follow the TEA guidelines and stubify tkinter, you
should not have any problems when upgrading to a new version. I know for a
fact that this worked for me when I upgraded from wish 8.1.x to 8.2.x: I
didn't need to recompile the Img extension, which was properly stubified.

L

-- 
Penguin Power!           Nothing I say reflects the views of my employer

Laurent Duperval                          mailto:laurent.duperval at cgi.ca
CGI - FWFM Project                                 Phone: (514) 350-3368





From pereira at research.att.com  Thu Nov 25 13:05:56 1999
From: pereira at research.att.com (Fernando Pereira)
Date: Thu, 25 Nov 1999 18:05:56 GMT
Subject: natural speech parser?
References: <81jf82$kil$1@nnrp1.deja.com>
Message-ID: <251119991305568565%pereira@research.att.com>

In article <81jf82$kil$1 at nnrp1.deja.com>, <moonseeker at my-deja.com>
wrote:

> Hi,
> 
> I'm looking for a command parser to parse sentences like:
> 
> enable KEY1 and write 'Hello World' to var2
> 
> I havn't done such a thing like parsers or compilers (which works with
> command trees) since yet and would welcome any suggestion.
This can be relatively easy or an open research project, depending on
what you really want. If what you want is a restricted, artificial
language that looks a bit like natural language, it's not too difficult
to build a simple parser based on an appropriate specification, along
the lines of

   <command> ::= (<clause> <conj>)* <clause>
   <clause> ::= <intransverb> | <transverb> <obj> |
      <ditransverb> <obj> <prep> <iobj>
   <conj> ::= "," | "and"
   <intransverb> ::= "help" | "stop" | ...
   <transverb> ::= "enable" | "disable" | ...
   <ditransverb> ::= "write" | "send" | "copy" | "store" | ...
   <obj> ::= <var> | <const>
   <const> ::= <string> | <number> | ...
   <iobj> ::= <var>
   <prep> ::= "to" | "into" | ...

augmented with appropriate semantic actions. There are several Python
tools around to create parsers, eg.
<http://www.csr.uvic.ca/~aycock/python/>.

However, a restricted language may be awkward to use without training,
since actual natural language is more variable than can be easily
captured by such a grammar. The situation is even more difficult if the
parser is intended to work on the output of speech recognition, for
several reasons: 1) spoken language is more variable than written, 2)
actual speech is full of disfluencies, and 3) speech recognizers make
lots of mistakes. Dealing with these problems is the subject of quite a
bit of current research, and building actual usable systems requires
limiting the scope of application to match the limitations of the
technology, and a lot of user testing.

-- F



From tottinge at concentric.net  Tue Nov 30 08:21:12 1999
From: tottinge at concentric.net (Tim Ottinger)
Date: 30 Nov 1999 08:21:12 EST
Subject: WinNT and python as text filter...
Message-ID: <3842f32d.30322451@news.concentric.net>

I'm working in WinNT with VIM (good tool) and wanted to write some
handy filters in python to help me with routine editing chores. 

I run my python command at the command line and get the output I want
printed nicely on my screen. However, if I pipe it to a file the file
comes out empty, if I pipe it to more I get nothing at all.

Does python know it's in a pipe and act differently? Does 'print' not
work right in NT? What's up?

My python is:

  PythonWin 1.5.2 (#0, Apr 13 1999, 10:51:12) [MSC 32 bit (Intel)] on
win32
  Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
  Portions Copyright 1994-1999 Mark Hammond 
Tim



From fredrik at pythonware.com  Mon Nov 22 10:16:29 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 22 Nov 1999 16:16:29 +0100
Subject: Zlib-error
References: <383966A1.F5E146E8@bibsyst.no>
Message-ID: <003a01bf34fc$908bd7b0$f29b12c2@secret.pythonware.com>

Thomas Weholt <thomas at bibsyst.no> wrote:
> data_to_store = pickle.dumps(object)
> data_to_store = zlib.compress(data)
> db = shelve.open('test','cwr')
> db[id] = data_to_store
> db.close()
> 
> and later :
> 
> data = zlib.decompress(pickle.loads(db[id]))
> 
> the error occurs.

maybe you should decompress the data
*before* unpickling it?  like this:

    data = pickle.loads(zlib.decompress(db[id]))

(btw, note that shelve pickles the string you
store in the database.  if you're compressing
all items in the database, you can save a few
more bytes by using anydbm instead of shelve).

</F>




From S.I.Reynolds at cs.bham.ac.uk  Wed Nov 24 09:09:13 1999
From: S.I.Reynolds at cs.bham.ac.uk (Stuart Reynolds)
Date: Wed, 24 Nov 1999 14:09:13 +0000
Subject: rlcompleter
References: <383A87F6.5D2F@cs.bham.ac.uk> <slrn83l3ka.ik.talon@coursen.ifu.net>
Message-ID: <383BF189.52C2@cs.bham.ac.uk>

Stephen Coursen wrote:
> 
> Well, I never got it working right with Control+tab, so instead I use
> meta-tab (alt-tab on PC's).
> 
> readline.parse_and_bind( 'm-tab: complete' )
> 

Works a treat.

Cheers,

Stu



From brunomadv at ciudad.com.ar  Thu Nov  4 15:43:25 1999
From: brunomadv at ciudad.com.ar (Bruno Mattarollo)
Date: Thu, 4 Nov 1999 17:43:25 -0300
Subject: Generating charts [Q]
Message-ID: <003201bf2705$3f287900$86152ac8@infosys.com.ar>

Hello.

    I need to generate Charts (GIF, JPEG or PNG) to display information that I retrieve from a database. 

    I downloaded and start looking at PIL and gdchart. I need to have control (more is best) over all the aspects of the graphic. We are working with graphic designers that want this images to match with the overall aspect of the site. So I need to be able to control from the background color up to the font size displayed. What's the best choice? I have little time to start doing research on how to use a specific tool... :( As almost all software developers, all our work is suposed to be ready yesterday ;)

    TIA

/B

Bruno Mattarollo <brunomadv at ciudad.com.ar>
--- Python Powered <http://www.python.org/psa/>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/19991104/4ff9f0bf/attachment.html>

From aa8vb at yahoo.com  Fri Nov  5 12:51:24 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Fri, 5 Nov 1999 12:51:24 -0500
Subject: eff-bot's The Standard Python Library
Message-ID: <19991105125124.A302546@vislab.epa.gov>

     I must say this book is quite good.  For those that monitor the Python
list (among other reasons) to learn about the large set of standard
libraries and to pick up new coding techniques, you may want to look into this.

     The new info and slick tricks per unit area is much higher than
drinking from the Python list daily, so it's more enjoyable to read (and I
consider the list fun to read anyway).  It highlights each piece with a
short code segment, and thankfully, doesn't bury the knowledge and tricks
in large case studies.  Well organized and formatted.  I'd recommend it.
(Haven't read it cover-to-cover yet of course, but I've browsed the rest.)

     I'd buy it for more in print, so yes IMO it's certainly worth $16.
Nice to have it on-line for browsing and searching too.

     http://www.pythonware.com/people/fredrik/librarybook.htm
     
     Hopefully eMatter will complete their efforts to implement a UNIX
distribution method soon.  But if you'd rather not wait, you're a UNIX
weanie, and have an MSWindows box available, "think creatively".  

     Download note: no trouble at all with the Fatbrain site.  I set my
account up as I purchased (one other person previously had reported
problems doing this).

     Install note: No luck running it on NT4; no GUI or console messages
appeared at all, though I didn't have Acroread installed there (possibly
what confused it, though I would still have expected an error message).  It
worked OK under W98 though.  Be sure to heed their advice and install
Acroread 4.x.  I rarely boot into MSWin so I still had Acroread 3.x
installed.  When given this PDF, 3.x was not a happy camper (missing
colormaps, invalid Postscript generation, etc.).  4.x worked like a champ
though.

-- 
Randall Hopper 
aa8vb at yahoo.com



From robin at alldunn.com  Fri Nov 19 14:47:23 1999
From: robin at alldunn.com (Robin Dunn)
Date: Fri, 19 Nov 1999 11:47:23 -0800
Subject: Building wxPython [was Re: wxPython and DISLIN work nicely together]
References: <80s0b4$g5n@mail.psy.uva.nl> <19991117090214.B807152@vislab.epa.gov> <017a01bf31f8$6bbea590$1a25d2d1@jenkondev.com> <19991119064944.A996886@vislab.epa.gov> <002301bf32b5$97a5d440$0301a8c0@easystreet.com> <19991119134422.A1004588@vislab.epa.gov>
Message-ID: <00e301bf32c6$e77aee90$0301a8c0@easystreet.com>

>
> And to the former, here's my 'wx-config --cflags' output:
>
>     > wx-config --cflags
>     -I/home/rhh/software/wxGTK-2.1.11/include -I/home/rhh/software/wxGTK-2
.1.11/lib/wx/include -D__WXMOTIF__ -I/usr/Motif-1.2/include
>
>

Well, this will be a problem for you if you get past he other compilation
issues as wxPython isn't ready for Motif yet.  (I'm going to attempt it
again soon.)  If you want to use wxPython now you'll need to build and
install glib and gtk version 1.2.3 (1.2.6 might work too) from www.gtk.org,
and then configure and build wxWindows with the --with-gtk flag.

> Leafing through build.py, the problem is that wxPython assumes wxGTK's
> wx-config is in $PATH.  I don't have root so I don't install things in
> /usr/local/*.  (I also don't like having package intermingle in a shared
> directory tree; too hard to remove them later.)

"make uninstall" works.

> So the change to hack
> around this is to temporarily add $WXWIN's bin directory to $PATH:
>
>      env PATH=$WXWIN/bin:$PATH /usr/local/bin/python \
>          $WXWIN/utils/wxPython-2.1.11/distrib/build.py -b MAKE=gmake
>
> Perhaps wxPython should have a --with-wxwindows= option, or take it's
> queues off of $WXWIN?
>

wxWindows is designed to use wx-config for all apps that are built using
wxWindows, (except for its samples/tests.)  That is why it made a bin
directory under your install prefix.  The gtk and glib libraries work this
same way.


> It then went on a little further to try and compile helpers.cpp, but
failed
> with 84 errors, all (or almost all) due to a bad NULL comparisons in the
> wxWindows header files.  NULL is defined as ((void*)0) and cannot be
> assigned to / compared with another type without a cast in C++.  As this
is
> C++, I believe they should be using "0" for assignments and comparisons.
>

I don't know how accurate this is but I've been told that C++ compilers that
have NULL defined as ((void*)0) in their headers are the ones that are
incorrect, not the code that uses NULL.  The headers are supposed to define
it as 0.  Maybe you could try undef'ing it and redefining NULL to be 0 and
see what happens?

This makes me wonder how you built wxWindows at all if these errors are in
the wxWindows headers...  What about the wxWindows sample apps, do they
compile?


--
Robin Dunn
Software Craftsman
robin at AllDunn.com
http://AllDunn.com/robin/
http://AllDunn.com/wxPython/  Check it out!





From skip at mojam.com  Fri Nov 12 10:56:40 1999
From: skip at mojam.com (Skip Montanaro)
Date: Fri, 12 Nov 1999 09:56:40 -0600 (CST)
Subject: '==' vs. 'is' behavior
In-Reply-To: <382C2C6A.64A8F2C2@theriver.com>
References: <382C2C6A.64A8F2C2@theriver.com>
Message-ID: <14380.14520.317089.290351@dolphin.mojam.com>

    Andy> Further experimentation shows that this seems to work for integers
    Andy> through 99, but that starting with 100 the results of the 'is'
    Andy> test return 0, as I would expect.

Python's intobject.c file maintains a cache of the integer objects 0 through
99.  If you check the id() of small integers you'll see they are always
represented by the same object.

read-the-source-luke-ly y'rs,

Skip Montanaro | http://www.mojam.com/
skip at mojam.com | http://www.musi-cal.com/
847-971-7098   | Python: Programming the way Guido indented...



From a_stepanov at yahoo.com  Sat Nov  6 22:53:16 1999
From: a_stepanov at yahoo.com (Alexander Stepanov)
Date: Sat, 06 Nov 1999 22:53:16 -0500
Subject: Python and generating images for web
Message-ID: <3824F7AC.4B1BC7F6@yahoo.com>

Hello,


I am a newcomer to web programming and python.  I need to create a small
web application
to make some charts based on user's data input.

Is it possible to generate gif or png images on a fly using Python?

I will appreciate if you point me out accessible chart/plot  modules and
libraries as well as documentation and
working sites.    Any additional information is strongly welcome.

(I am in a process of comparison Perl and Python to solve this problem.
I plan to use the application
under WindowsNT4.0 environment + Perl? or Python interp-r).

Sincerely,  Alexander
email: a_stepanov at yahoo.com





From gvwilson at nevex.com  Thu Nov 25 14:10:21 1999
From: gvwilson at nevex.com (Greg Wilson)
Date: Thu, 25 Nov 1999 14:10:21 -0500
Subject: XML to in-memory tree?
Message-ID: <tRf%3.8219$cz3.293110@nnrp1.uunet.ca>

Hi.  Has anyone built a layer on top of XMLLIB that will load an XML
document into an in-memory (similar to the one built by explicit calls to
HTMLgen)?
Thanks,
Greg






From rshory at nrel.colostate.edu  Wed Nov 17 11:35:25 1999
From: rshory at nrel.colostate.edu (Rick Shory)
Date: Wed, 17 Nov 1999 09:35:25 -0700
Subject: Python in Windows CE?
Message-ID: <3832D94D.50A25475@nrel.colostate.edu>

  Has anybody used Python for development under Windows CE?
-- 
rshory at nrel.colostate.edu
Natural Resource Ecology Laboratory A225
Colorado State University, Fort Collins CO 80523-1499
970-491-5835, fax 970-491-1965



From cjc26 at nospam.cornell.edu  Thu Nov 25 23:27:46 1999
From: cjc26 at nospam.cornell.edu (Cliff Crawford)
Date: 26 Nov 1999 04:27:46 GMT
Subject: The First Annual Obfuscated Python Contest
References: <mfbZ3.71$AZ6.1807@dfiatx1-snr1.gtei.net> <001d01bf328d$3ac77170$f29b12c2@secret.pythonware.com>
Message-ID: <slrn83s2vv.1s0t.cjc26@tankgrrl.bridget.mindriot.net>

Pada Fri, 19 Nov 1999 13:54:32 +0100, Fredrik Lundh bilang:
| 
| for more masterpieces, see:
| 
|     http://starship.python.net/crew/amk/writing/crypto-curiosa.html
|     (from where the slashdot item was stolen)
| 
|     http://www.python.org/doc/FAQ.html#4.15
| 
| and greg's all-time masterpiece:
| 
| http://www.deja.com/getdoc.xp?AN=398922825&fmt=text
| (make sure to place it all on one line)

Notice that all examples of obfuscated Python make liberal use of `lambda',
`map', and `reduce'.

next-they'll-want-obfuscated-Haskell-ly yr's, Cliff


-- 
cliff crawford   http://www.people.cornell.edu/pages/cjc26/
-><-                          Java is the COBOL of the 90s.



From nick at src.uchicago.edu  Wed Nov  3 15:45:50 1999
From: nick at src.uchicago.edu (Nick Collier)
Date: Wed, 03 Nov 1999 14:45:50 -0600
Subject: Can't start IDLE
References: <381F4EB1.AFD9BA32@src.uchicago.edu> <5lg0yotka6.fsf@eric.cnri.reston.va.us> <38209D52.C4370040@src.uchicago.edu>
Message-ID: <38209EFE.271AD282@src.uchicago.edu>

Just tried 

import Tkinter

Tkinter._test() and it works, so maybe its not a tk version problem.

Nick

Nick Collier wrote:
> 
> It looks like this is the problem. Unfortunately, some other software
> I've got installed relies on other versions of tcl, so I'll have to run
> emacs as my python editor for now. Saw IDLE at the talk you gave at the
> Univ. of Chicago last week and wanted to give it a try. I enjoyed the
> talk and thanks for coming.
> 
> Nick
> 
> Guido van Rossum wrote:
> >
> > Nick Collier <nick at src.uchicago.edu> writes:
> >
> > > On winNT sp3 I get the following message when I do a "python idle.py"
> > > from the command line:
> > >
> > > C:\Program Files\Python\Tools\idle>python idle.py
> > > Traceback (innermost last):
> > >   File "idle.py", line 3, in ?
> > >     PyShell.main()
> > >   File "PyShell.py", line 612, in main
> > >     fixwordbreaks(root)
> > >   File "EditorWindow.py", line 606, in fixwordbreaks
> > >     tk.call('tcl_wordBreakAfter', 'a b', 0) # make sure word.tcl is
> > > loaded
> > > TclError: invalid command name "tcl_wordBreakAfter"
> > >
> > > I looked around deja.com and found one other person with the same
> > > problem, but no response so I'm asking here.
> >
> > Could it be that you have an older (or newer) Tcl version installed?
> > Python (for now, until I upgrade the _tkinter module) requires Tcl/Tk
> > 8.0.5 -- which is the version that comes with the Python 1.5.2 for
> > Windows installer.
> >
> > Try to erase all traces of other Tcl installs and reinstall Python and
> > see if that works.
> >
> > --Guido van Rossum (home page: http://www.python.org/~guido/)
> 
> --
> 
> Nick Collier
> nick at src.uchicago.edu
> Social Science Research Computing
> University of Chicago
> Chicago, IL 60637

-- 

Nick Collier
nick at src.uchicago.edu
Social Science Research Computing
University of Chicago
Chicago, IL 60637



From init__self at my-deja.com  Mon Nov  8 13:57:19 1999
From: init__self at my-deja.com (init__self at my-deja.com)
Date: Mon, 08 Nov 1999 18:57:19 GMT
Subject: pythonistas or pythoneers?
References: <Pine.LNX.4.10.9911081026470.4450-100000@robert.polytopic.com> <slrn82e4mp.43f.thor@localhost.localdomain>
Message-ID: <8076ed$d2n$1@nnrp1.deja.com>

> So pythonistas sounds extremely powerfull, because python can
> be spelled in Spanish as "piton", that means "horn" ( like the bull's
> horns). So pythonistas sounds like the "damned crazy followers
> of the horns" :DD

So instead of those masturbatory "Just another you-know-what hacker"
.sig's, one could have

   #!/usr/local/bin/python
   print "Just another damn crazy follower of the horns."

as a fine .sig :-) It conveys succinctly the readability and simplicity
of the language!


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



From banderson at boi.hp.com  Tue Nov 16 13:39:39 1999
From: banderson at boi.hp.com (Bill Anderson)
Date: Tue, 16 Nov 1999 18:39:39 +0000
Subject: Beginning programming with Python
References: <sng0ychgwt.fsf@motorola.com> <382B3A53.5816B2@callware.com> <382B3D79.74CD3E0F@email.sps.mot.com> <80p9ds$ppo$1@newshost.accu.uu.nl> <asteroidal-80psfo/INN-2.2.1/bat@broadway.news.is-europe.net> <38312D55.A36EB56D@iqsoft.hu>
Message-ID: <3831A4EB.76FF32AF@boi.hp.com>

Nemeth Miklos wrote:
> 
> Janos Blazi wrote:
> 
> > I have tried to use Perl in my teaching. It has a very powerful but alas!
> > not very simple syntax! To start with, my pupils used to forget the dollar
> > signs and Perl does not like that. So this year I shall use Python. We shall
> > see. Time will tell, if Python is really better. Maybe my pupils will forget
> > the indenting this time? There is actually nothing else you can forget in
> > Python.
> 
> I am quite new in Python programming, and I was a bit :-( by understanding that
> Python does not have any (even optional) mechanism to enforce the usage of
> predeclared variables and functions. The complete lack of compile-time
> type-safety of Pythom may be a problem in large projects, which may be the most
> serious technical (ie non-marketing) obstacle to Python's becoming a widespread
> language like Java.
> Let us see an example:
> # Module t.py
> def f():
>   print "started"
>   x = 128
>   x.foo()
> 
> Let us import it in the Python interpreter:
> >>> import t
> >>>
> You can see thet the interpreter compiled t.py to t.pyc (or t.pyo) but found no
> error.
> All typos (typing errors of variable and function names) will be detected ONLY
> at runtime! Python is perfect language to write programs fast, but with the
> cost of hordes of testers.

Disagree. I write many things in python, fast, and without hordes of
testers.

> However this may not be a big problem, because any
> piece of software should be thoroughly tested, to make sure that there are no
> semantic errors in the application. In the case of Python the applications
> should also be tested for "syntactic" errors.

As should _any_ language.

> In Python (just like in Smalltalk or in Perl, or in Tcl) there are a lot of
> things a programmer may forget.

Just like in C, java, C++ ... 

> I am on my way of becomming a Python programmer, but this should not prevent me
> from seeing the pitfalls of the language.
> A tutor must emphasize these gotchas of any programming language.
> Perl has a facility called 'use strict': it is also very poor compared to C++
> and Java strict static (ie compile time) system.
> The only really large project implemented in Python I found is Zope. Zope is a
> fantastic piece of software, and may be regarded as an evidence of my being
> totally wrong ;-).

yup :-)

> > I feel that it is the right language for teaching.
> 
> Anyway, Python is a great language for teaching, but why to learn a language,
> which is good only for teachning?! This is the real question you (or the Python
> community) should answer!

Not possible. The question is predicated upon an assumption. You
incorrectly assume that Python is _only_ good for teaching. There are
quite a number or large companies that would disagree with you.

Bill
 

-- 
Bill Anderson 				Linux/Unix Administrator, Security Analyst
ESBU (ARC)				banderson at boi.hp.com
My opinions are just that; _my_ opinions.



From mlh at vier.idi.ntnu.no  Sat Nov 13 11:07:28 1999
From: mlh at vier.idi.ntnu.no (Magnus L. Hetland)
Date: 13 Nov 1999 17:07:28 +0100
Subject: Random text generator
References: <199911120349.OAA01177@envy.fulcrum.com.au> <y0jk8nntzvn.fsf@vier.idi.ntnu.no> <01e101bf2dcd$f4bb6290$f29b12c2@secret.pythonware.com>
Message-ID: <y0jbt8ytlxb.fsf@vier.idi.ntnu.no>

"Fredrik Lundh" <fredrik at pythonware.com> writes:

> Magnus L. Hetland <mlh at vier.idi.ntnu.no> wrote:
> > >    The function you want is whrandom.choice(). See
> > 
> > Or - rather, random.choice() (which uses whrandom.choice, if
> > available)... No?
> 
> it would probably be a bit clearer (at least to newcomers)
> if "random" is was the official interface, and whrandom was
> treated as an implementation module.

I agree.

> but it doesn't look
> that way in the current version of the library reference...

No... But that's how it *is*, though - right?

>>> import random
>>> random.random
<method whrandom.random of whrandom instance at f2b48>
>>> 

I-hate-those-names-with-strange-letter-prefixes'ly yours,

Magnus

--

  Magnus          Echelon jamming noise:
  Lie             FBI CIA NSA Handgun Assault Bomb Drug Terrorism
  Hetland         Special Forces Delta Force AK47 Hillary Clinton 



From emmanuel.pietriga at xrce.xerox.com  Fri Nov 26 04:30:56 1999
From: emmanuel.pietriga at xrce.xerox.com (Emmanuel Pietriga)
Date: Fri, 26 Nov 1999 10:30:56 +0100
Subject: XPython (X11)
Message-ID: <383E5350.1E07536A@xrce.xerox.com>

Is anyone using XPython around here?

I need help, or at least links to examples (I haven't found any but the
ones within the package)

Thanks.
Emmanuel




From tim_one at email.msn.com  Wed Nov 24 01:32:19 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 24 Nov 1999 01:32:19 -0500
Subject: whrandom seed question
In-Reply-To: <81eu22$411al$1@umbc7.umbc.edu>
Message-ID: <000601bf3645$a8dc0ca0$f82d153f@tim>

[Stelios Kyriacou]
> I tried today the random number generator in python
> and have the following question:

Don't use module whrandom.  Use module random.  The former is an
implementation detail of the latter, and follows the fine scientific
tradition of using incomprehensible calling conventions <0.5 wink>; whrandom
isn't meant to be used directly; random is; someday the docs will catch up
with reality.

> Shouldn't the following 3 lines produce the same number?
>
> >>> whrandom.seed(0,0,0); print whrandom.random()
> 0.39613674715
> >>> whrandom.seed(0,0,0); print whrandom.random()
> 0.170188890936
> >>> whrandom.seed(0,0,0); print whrandom.random()
> 0.831213988577
> >>>
> ...

That's functioning as designed.  You want random.seed and random.random:

>>> import random
>>> random.seed(0); print random.random()
0.0169309061997
>>> random.seed(0); print random.random()
0.0169309061997
>>> random.seed(0); print random.random()
0.0169309061997
>>> print random.seed.__doc__
Seed the default generator from any hashable value.

	None or no argument returns (0, 0, 0) to seed from current time.


>>>

The last sentence there is damaged.  The docstring for random.seed should be

	"""Seed the default generator from any hashable value.

	No argument, or argument None, seeds from current time.
	"""

state-of-sin-indeed-ly y'rs  - tim





From a.eyre at optichrome.com  Thu Nov 25 11:14:01 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Thu, 25 Nov 1999 16:14:01 -0000
Subject: decompress a bunch of .gz files
In-Reply-To: <14397.18620.87516.687622@gargle.gargle.HOWL>
Message-ID: <000601bf3760$1649d9f0$3acbd9c2@peridot.optichrome.com>

>> #!/bin/sh
>> for file in *.gz; do gunzip $file; done

> gunzip *.gz

True. I just remebered you could do that after I posted.
Some Unix proglets don't let you do that.

My point was that Python is overkill for gunzip'ping files.

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From robin at jessikat.demon.co.uk  Tue Nov  9 12:25:26 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Tue, 9 Nov 1999 17:25:26 +0000
Subject: Graphics.
References: <3826E949.BBA1BAFD@imf.au.dk>
 <091119991052098844%kantel@mpiwg-berlin.mpg.de>
 <80973f$poq@newton.cc.rl.ac.uk> <m3zownmz6t.fsf@greebo.nodomain.de>
Message-ID: <kQsE5GAGkFK4EwcH@jessikat.demon.co.uk>

In article <m3zownmz6t.fsf at greebo.nodomain.de>, Bernhard Herzog
<herzog at online.de> writes
>"Richard Brodie" <R.Brodie at rl.ac.uk> writes:
>
>> ...and as everyone knows, if you haven't done Python in German, you
>> haven't done Python :)
>> 
>> Ich bin ein Holzf?ller und f?hl mich stark,
>> Ich schlaf des Nachts und hack am Tag.
>
>Shouldn't that have been
>
>Ich bin ein Pyonier und f?hl mich stark,
>Ich hack des Nachts und schlaf am Tag.
>
>?
>
>
I'm a lumberjack and I'm OK, usw usw
-- 
Robin Becker



From fredrik at pythonware.com  Tue Nov  2 04:24:02 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 2 Nov 1999 10:24:02 +0100
Subject: About RedHat 6.1 install program
References: <7vm92d$i1e$1@nnrp1.deja.com>
Message-ID: <000701bf2514$01458490$f29b12c2@secret.pythonware.com>

johndoe0521 at my-deja.com wrote:
> I think this might be python's problem.

have you contacted redhat?  did they tell you
to ask here?

if not, try:

http://www.redhat.com/cgi-bin/support_switch

if they *did* tell you to ask here, you most provide
more information.  what did they say?  what errors
do you get?  what exactly are you doing?  etc.

</F>




From madq968 at djeksta.comNOSPAM  Sat Nov  6 12:34:23 1999
From: madq968 at djeksta.comNOSPAM (madQ)
Date: 6 Nov 1999 17:34:23 GMT
Subject: Download Ia.n.i.!!! It's free!
Message-ID: <801oqv$9qr$407@nslave1.tin.it>

Download Ia.n.i. RemoteControlSystem 1.2 beta. It's free!!!
New site: http://members.tripod.de/madq/




From francois at roya.org  Tue Nov 16 13:42:46 1999
From: francois at roya.org (François Desvallées)
Date: Tue, 16 Nov 1999 19:42:46 +0100
Subject: wxPython on Solaris??
Message-ID: <942778029.608457@jerusalem.magic.fr>

Hello:

I need to to display graphics (raw bitmaps) on a Solaris workstation. I am
thinking of using wxPython, but I have only seen reports of it being used on
Windows and Linux. Has this been done, and how much work would it require
(using the Motif interface, I guess?)

Thanks
Francois





From g_will at cyberus.ca  Mon Nov  8 10:52:39 1999
From: g_will at cyberus.ca (Gordon Williams)
Date: Mon, 08 Nov 1999 15:52:39 GMT
Subject: Windows Sound Card
Message-ID: <01bf2a01$27e7cda0$8355c3d1@Cyberus.cyberus.ca>

Hi,

I am looking for information on a module to get data from the input from a
sound card on a Windows box.  Is this possible using Python??  I have had a
search through all the docs and FAQs but havent found anything for Windows.

Any pointers would be a help.

Thanks,

Gordon Williams



From curtis_c at lxe.com  Fri Nov 19 10:59:47 1999
From: curtis_c at lxe.com (Curtis, Craig M.)
Date: Fri, 19 Nov 1999 10:59:47 -0500
Subject: Python in DOS
Message-ID: <DB6390A9770ED211B2440000F807FC7B01039D69@nt_exlxe.lxe.com>

I tried setting PYTHONCASEOK to 0 and did the configure and make and still
have the same problem.  Any more suggestions?

Thanks,
Craig

> ----------
> From: 	Hans Nowak[SMTP:hnowak at cuci.nl]
> Sent: 	Friday, November 19, 1999 1:15 AM
> To: 	python-list at python.org
> Cc: 	parkw at better.net; Curtis, Craig M.
> Subject: 	Re: Python in DOS
> 
> 
> On 18 Nov 99, William Park wrote:
> 
> > On Thu, Nov 18, 1999 at 06:24:25PM -0500, Curtis, Craig M. wrote:
> > > Does anyone know the "approved" way to get rid of the following
> > > message:
> > > 
> > > Could not find platform dependent libraries <exec_prefix>
> > > Consider setting $PYTHONHOME to <prefix>[:<exec_prefix>]
> > > 'import site' failed; use -v for traceback
> > > Python 1.5.2 (#6, Nov 18 1999, 17:19:05)  [GCC 2.8.1] on ms-dos7
> > > Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> > > 
> > > 
> > > I've tried everything I can think of to set PYTHONHOME and PYTHONPATH
> > > and python ALWAYS starts up this way.
> > 
> > If I remember (this was long long time ago), I had to remove case
> > sensitivity for library files by setting some dummy environment
> > variable.
> 
> You surely mean the PYTHONCASEOK variable. If that doesn't help, 
> simply comment these parts out of the source code... DOS is not going 
> to use them anyway.
> 
> 
> --Hans Nowak (zephyrfalcon at hvision.nl)
> Homepage: http://fly.to/zephyrfalcon
> You call me a masterless man. You are wrong. I am my own master.
> 



From guido at CNRI.Reston.VA.US  Thu Nov  4 08:25:53 1999
From: guido at CNRI.Reston.VA.US (Guido van Rossum)
Date: Thu, 04 Nov 1999 08:25:53 -0500
Subject: __str__ vs. __repr__
In-Reply-To: Your message of "Wed, 03 Nov 1999 20:58:22 PST."
             <Pine.WNT.4.05.9911032048030.109-100000@david.ski.org> 
References: <Pine.WNT.4.05.9911032048030.109-100000@david.ski.org> 
Message-ID: <199911041325.IAA14001@eric.cnri.reston.va.us>

[David Ascher]
> I certainly appreciate the justification for the current behavior, which I
> didn't really understand either.
> 
> I have two suggestions which would make the current read-eval-print
> behavior somewhat more palatable to me:
> 
>  -- container types (lists, dicts, tuples) should have both repr and str
>     methods, even if they differ only in the methods of the contents that
>     they call.  You can make the str methods use the repr-type display
>     (with brackets/parens/braces and commas), as that wouldn't break
>     anyone's expectations (since that's what is currently being used), and
>     would 'solve' the current inconsistency.
> 
>     I understand str output is meant to be human readable -- that
>     doesn't mean that []'s and ,'s can't be used.

I'm reluctant but Tim's dictionary-of-rationals example suggests
you're right.

>  -- make the str of a long integer not have the postfix L.  After all, if
>     the __str__ is supposed to be for human consumption, the L is no more
>     useful than the quotes around strings.  

Good one!

> The issue of determining the printability of characters is one *I*
> wouldn't touch with a 10-foot snake.

Unfortunately I will have to touch this...  I cannot think of
something better than isprint(), which is designed for this.  But the
weirdness is that if I write a tiny test program, the default (== C
locale) behavior is for isprint() to return false for anything over
\177; however in Python, it seems to always assume a Latin character
set.  To be precise, the following non-ASCII characters are considered
printable: '?????????????????????????'.  Is there anything I may have
forgotten about isprint()?  Using the Python locale module's
setlocale() function doesn't seem to make a difference, which is also
weird (Python doesn't call setlocale(), so it *should* default to the
C locale.  (And yes, I'm including <ctype.h>.)

I have one more suggestion myself: add an "expert mode" where values
are displayed using str() instead of repr() by the read-eval-print
loop.  I call it expert mode because it's only useful when (a) you are
manipulating custom types like Tim's rational number class, and (b)
you are aware of the ambiguities and risks of using str() (garbage
characters and multiple interpretations).

By the way, there's another issue here that fortunately hasn't muddled
the discussion.  The C API allows an object to specify *three* ways of
formatting itself: str(), repr() and print-to-file.  The print-to-file
call should faithfully mimic either str() or repr() depending on a
'mode' argument flag; however some 3rd party extensions (none in the
core distribution as far as I know) have a str() or repr()
implementation that deviates from print-to-file.  These should be
fixed.  Forget I even mentioned this in this thread, though!

--Guido van Rossum (home page: http://www.python.org/~guido/)



From irwin at mail.com  Wed Nov 17 09:14:06 1999
From: irwin at mail.com (Roger Irwin)
Date: Wed, 17 Nov 1999 15:14:06 +0100
Subject: ASCII delimited....a follow up
Message-ID: <99111716290505.00284@snoopy>

Thanks to everybody who sent me suggestions for reading delimited ASCII files
last week. As there was no standard funtion for doing what I wanted I wrote my
own, but as this could be very useful to a lot of other people (if nothing alse
for reading in data from spreadsheets), I thought others might like a copy of
this simple function:

(Note, this is more orientated towards the requirements of general data tables
rather than mathematical data sets).

pdict={}

def parsedata(ftr,delim=';',offset=0,strip=1):
	"Read a delimited ASCII file into a dictionary"

	"After opening the file and reading in any header lines,"
	"this function can be used with ftr= to the open file object"
	"The delim parameter is obvious whilst the offset points to"
	"the field to be used as the key."
	"If strip is true, then if a field starts with a quotation"
	"the first and last character will ripped off the field."
	"A line with no delimiter will be regarded as a comment"
	"Stops reading as soon as a control character is encountered"

	"Returns number of records read"
	
	counter=0
	while 1:
		foo=ftr.readline()
		if foo=='':
			break		
		start=0
		buf=[]
		parsetab=[]

		for each in range(len(foo)):
			if foo[each]==delim:
				if (foo[start]=='"')|(foo[start]=="'"):
					parsetab.append(foo[start+1:each-1])
				else:
					parsetab.append(foo[start:each])
				start=each+1
			elif foo[each]<' ':
				if start:
					if (foo[start]=='"')|(foo[start]=="'"):
						parsetab.append(foo[start+1:each-1])
					else:
						parsetab.append(foo[start:each])
				break
		
		if start:
			pdict[parsetab[offset]]=parsetab	
			counter=counter+1
	return counter
	

Note that all fields get read as strings. If you need to do sums on a value
then it is better (IMHO) to do a specific conversion dependent on requirements
after the file is read in, like using getline/sscanf in C. Many programs (well,
mostly spreadsheets) put quotation marks round strings in deliminted files
whilst numbers are left as is. These quotation marks will get included in the
string when reading into Python. I thought about doing something
hyper-programmable to allow for all possibilities of removing these quoteation
marks, but then I decided this was overkill, took the KISS option, and did a
simple test whereby if the first character of a field is a quotation mark then
the first and last character will be stripped. This will cater for virtually
all real world requirments, but it is overidable in awkward cases.

Now I never wanted to be a programmer, I always wanted to be a lumberjack, but
instead I am a newbie Pythonite. As such I would appreciate feedback on the
way I have done this, particularly the way I have used pdict{}.

I wanted to have the possibility of reading several files into one dictionary,
whatsmore the files may have header records, so rather than have the parser
actually open and close the file, that is done by the calling
function, which may also read in any header records before passing
the file object pointer to the function. 

Now, if I had been in C, I would probably have passed a pointer to an array of
structures, but Python does not seem to encourage passing by reference in
function calls (or have I missed something), so I used a specific temporary
dictionary location. This way I can open several files in succession and
concatanate their contents into the same dictionary. If I then want to start a
new dataset (and this code were in 'module'), I would do:

firstdataset=module.pdict
module.pdict={}

And thus carry on reading a new lot of data. Now I assume that Python would not
have copied the disctionary, but simply passed a pointer to firstdataset, and
then created a new data object for module.pdict when I made the assignment. Is
this correct? Is there a better way to do this?

BTW, python would seem to be a great language for spreadsheets. Are there
spreadsheets with Python bindings, or perhaps even an extensible spreadsheet
written in Python?




From fredrik at pythonware.com  Tue Nov 16 04:59:32 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 16 Nov 1999 10:59:32 +0100
Subject: list.without()?
References: <NDBBIAPHEKGOBIHCBLIICEOMCAAA.mcfletch@vrtelecom.com> <y0jiu32iyi8.fsf@vier.idi.ntnu.no>
Message-ID: <01b001bf3019$48c5a010$f29b12c2@secret.pythonware.com>

Magnus L. Hetland <mlh at vier.idi.ntnu.no> wrote:
> Why on *earth* would you use a while/try/break-combination?

maybe because the meaning of "without a
specified element" is unclear.  mike's version
guarantees that:

    list.without(element)
    assert element not in list

never fails, while your version behaves just
like "remove" -- if there are several instances
of an element value, only the first one is
removed:

    count = list.count(element)
    list.without(element)
    assert count == 0 or list.count(element) == count-1

in either case, adding "without" (like remove, but
returns a new list) strikes me as being even more
contrived than earlier proposals like "sorted" (like
sort, but returns a new list), etc.

</F>




From endgamer71 at hotmail.com  Tue Nov 16 12:18:19 1999
From: endgamer71 at hotmail.com (Endgamer)
Date: Tue, 16 Nov 1999 17:18:19 GMT
Subject: Alright, I give up...chalk this one up to experience...
References: <01BF303A.7AFA1770@exchange1.geometria.hu>
Message-ID: <383190c4.18459747@212.225.36.99>

On Tue, 16 Nov 1999 13:57:10 +0100, "Kiss, Arpad" <AKiss at GEOMETRIA.hu>
wrote:

>Hi,
>
>Maybe date is waiting for your new date setting, but why don't you use 
>time instead of it:
>
>import time
>print time.ctime(time.time())
>print time.localtime(time.time())
>
>Arpad Kiss

Thanks Arpad, and all you guys. And yes, I know, I completely fucked
up the bit I posted (as I say I've been bashing on this for a long
time, brain starting to give up) - left in the test line. Maybe I'm
not cut out for learning Python...


Endgamer
Spider Warrior

(Endgamer71 at spamtrap.hotmail.com)
#####################################
Not a quote, just an on-the-spot observation:
You know you're a geek when your PC costs exactly
15 times as much as your stereo. And gives better
sound quality.<sigh>
                              Endgamer
#####################################



From drjatalb at my-deja.com  Wed Nov 24 13:23:25 1999
From: drjatalb at my-deja.com (drjatalb at my-deja.com)
Date: Wed, 24 Nov 1999 18:23:25 GMT
Subject: creating an AIX shared object for Python
Message-ID: <81haen$5qg$1@nnrp1.deja.com>

My question is in trying to create a shared object that can be
dynamically loaded into Python on AIX.  I have a small extension module
of C code which makes a 3rd party C library available to Python.  I
have been successful in creating a shared object on IRIX, Solaris, HP-
UX, and Windows NT for this extension module but so far have not been
successful in doing it under AIX.  Here are the two simple commands I
am trying to use and the error message which results during the build
of the shared library (I am using Python 1.5.2):

cc -c -qfuncsect -I/users/toollib/include/python1.5 -
I/net/rush/disk1/pdi/toolkits/crusher/crusher312/include crushermodule.c

ld -G crushermodule.o -o crushermodule.so -
L/net/rush/disk1/pdi/toolkits/crusher/crusher312/lib/AIX_4.2.1 -lcrusher
ld: 0711-327 WARNING: Entry point not found: __start
ld: 0711-244 ERROR: No csects or exported symbols have been saved.

Any suggestions would be appreciated.

Dr. Jeff Talbert
jat at iti-oh.com


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



From a.eyre at optichrome.com  Tue Nov 23 11:51:02 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Tue, 23 Nov 1999 16:51:02 -0000
Subject: Python and COM: a bug?
In-Reply-To: <816010E2456BD111A48700805FBBE2EEFDF796@ex-quebec-u1.baan.com>
Message-ID: <001d01bf35d2$ecbd0c80$3acbd9c2@peridot.optichrome.com>

> I have encountered a weird problem a few times in the past months. I'm
> testing COM server (written here) with Python, and it seems that there is
> something wrong in the code that optimizes the calls to COM objects.
> Sometimes (a very rare occurence) the wrong function gets called when I
> access a property.

We have had similar problems here with PythonCOM, and have had to stop
using it, finding its behaviour too random.

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From parkw at better.net  Thu Nov 11 12:23:20 1999
From: parkw at better.net (William Park)
Date: Thu, 11 Nov 1999 12:23:20 -0500
Subject: .html -> .ps ?
In-Reply-To: <14378.63638.48886.564135@weyr.cnri.reston.va.us>
References: <19991111080506.A1007@better.net> <14378.54900.714487.334690@weyr.cnri.reston.va.us> <19991111120757.C1783@better.net> <14378.63638.48886.564135@weyr.cnri.reston.va.us>
Message-ID: <19991111122320.A1862@better.net>

On Thu, Nov 11, 1999 at 12:10:46PM -0500, Fred L. Drake, Jr. wrote:
> 
> William Park writes:
>  > Found the solution by doing 'netscape -help'... duh!  See
>  >     http://home.netscape.com/newsref/std/x-remote.html
>  > Basically,
>  >     $ netscape -remote "openURL(http://...)"
>  >     $ netscape -remote "saveAs(test.ps,Postscript)"
> 
>   Ah, so you don't mind having a window opened on the screen.  I
> figured you were trying to get around that as well.

I think I can use '-noraise' to prevent the window opening... Yes, just
tried it.

William Park



From s.havemann at tu-bs.de  Wed Nov 17 13:42:03 1999
From: s.havemann at tu-bs.de (Sven Havemann)
Date: Wed, 17 Nov 1999 19:42:03 +0100
Subject: Embedded Python 1.5.1 crashes under WIN32
Message-ID: <3832F6FB.68AAA96E@tu-bs.de>

Hi!
===

I've modified the example from Python-1.5.1/Demo/embed: 

ftp://ftp.cg.cs.tu-bs.de/pub/cg/people/havemann/embeddedPython.tgz

What I want to do is to call PyRun_SimpleFile from my main program. This
works fine under Irix 6.5, but the same thing crashes inside
PyRun_SimpleFile when compiled with VC6.0SP2 under Win32, which is a
great pity. I have attached the demo.cpp file, and the complete
(compiled) project is in the tarfile above. Does anybody know why it
crashes??? 

I wanted to build it as a plain Win32 console application, and I've
renamed it to demo.cpp because I wanted it to be compiled as a C++ file. 

I'd appreciate if you could Re: to me directly using the email below,
because I can't follow the news regularly.

TIA, greetings, Sven.

--
__________________________________________________________________
 dipl-inform. Sven Havemann        Institut fuer ComputerGraphik
 Odastrasse 6                      Rebenring 18
 38122 Braunschweig - Germany      38106 Braunschweig - Germany
 Tel. 0531/2808955                 Tel. 0531/391-2108, Fax: -2103
 mailto:s.havemann at tu-bs.de        http://www.cg.cs.tu-bs.de


=========================demo.cpp==================================
#include <stdio.h>
#include "Python.h"

// #define SHOWPYFILE

static char *argv0;

void initxyzzy(); /* Forward */

int main(int argc, char ** argv)
{
   argv0 = argv[0];
   Py_Initialize();

   initxyzzy();

   PySys_SetArgv(argc, argv);

   printf("Hello, brave new world\n\n");

   PyRun_SimpleString("import sys\n");
   PyRun_SimpleString("print sys.builtin_module_names\n");
   PyRun_SimpleString("print sys.modules.keys()\n");
   PyRun_SimpleString("print sys.argv\n");

   char* filename="test.py";
   FILE* myfile=fopen(filename,"r");
#ifdef SHOWPYFILE
   char line[200];
   printf("Reading file %s\n",filename);
   while(fgets(line,200,myfile)!=NULL) {
      printf("Read: %s",line);
   }
#else
   PyRun_SimpleFile(myfile,filename);
#endif
   fclose(myfile);

   printf("\nGoodbye, cruel world\n");

   Py_Exit(0);
   return 0;
}

char *getprogramname()
{
        return argv0;
}

/* A static module */

PyObject *xyzzy_foo(PyObject *self, PyObject *args)
{
        if (!PyArg_ParseTuple(args, ""))
                return NULL;
        return PyInt_FromLong(42L);
}

PyMethodDef xyzzy_methods[] = {
        {"foo",         xyzzy_foo,      1},
        {NULL,          NULL}           /* sentinel */
};

void initxyzzy()
{
        PyImport_AddModule("xyzzy");
        Py_InitModule("xyzzy", xyzzy_methods);
}
==================test.py======================================
print "this is from PyRun_SimpleFile"
for i in range(100): print i*i,
==================OUTPUT IRIX==================================
Hello, brave new world

('__builtin__', '__main__', 'array', 'fast_umath', 'imp', 'marshal',
'math', 'operator', 'posix', 'signal', 'strop', 'struct', 'sys',
'umath')
['xyzzy', 'os', 'exceptions', '__main__', 'posix', 'sys', '__builtin__',
'site', 'signal', 'UserDict', 'posixpath', 'stat']
['./demo']
this is from PyRun_SimpleFile
0 1 4 9 16 25 36 49 64 81 100 121 144 169 196 225 256 289 324 361 400
441 484 529 576 625 676 729 784 841 900 961 1024 1089 1156 1225 1296
1369 1444 1521 1600 1681 1764 1849 1936 2025 2116 2209 2304 2401 2500
2601 2704 2809 2916 3025 3136 3249 3364 3481 3600 3721 3844 3969 4096
4225 4356 4489 4624 4761 4900 5041 5184 5329 5476 5625 5776 5929 6084
6241 6400 6561 6724 6889 7056 7225 7396 7569 7744 7921 8100 8281 8464
8649 8836 9025 9216 9409 9604 9801

Goodbye, cruel world
==================OUTPUT WIN32==================================
-- Hello, brave new world

('__builtin__', '__main__', '_socket', 'array', 'binascii', 'cPickle',
'cStringI
O', 'cmath', 'errno', 'imageop', 'imp', 'marshal', 'math', 'md5',
'msvcrt', 'new
', 'nt', 'operator', 'pcre', 'regex', 'rgbimg', 'rotor', 'select',
'signal', 'so
undex', 'strop', 'struct', 'sys', 'thread', 'time')
['xyzzy', 'os', 'exceptions', '__main__', 'ntpath', 'strop', 'nt',
'sys', '__bui
ltin__', 'site', 'signal', 'UserDict', 'string', 'stat']
['T:\\WORK\\HAVEMANN\\SW\\PYTHON-1.5.1\\DEMO\\EMBED\\VC60.exe']
=====================** CRASH **================================
================================================================
__________________________________________________________________
 dipl-inform. Sven Havemann        Institut fuer ComputerGraphik
 Odastrasse 6                      Rebenring 18
 38122 Braunschweig - Germany      38106 Braunschweig - Germany
 Tel. 0531/2808955                 Tel. 0531/391-2108, Fax: -2103
 mailto:s.havemann at tu-bs.de        http://www.cg.cs.tu-bs.de



From luser at no.where  Thu Nov  4 10:11:18 1999
From: luser at no.where (Linux User)
Date: Thu, 4 Nov 1999 10:11:18 -0500 (EST)
Subject: exceptions from C
Message-ID: <Pine.LNX.4.10.9911040950440.25629-100000@crb.crb-web.com>

I have been reading the section in the Lutz book on exceptions.  It seems
that exceptions are no longer simply string objects.  Does this change the way
exception are raised from C?  If so where might I find info on how to do this?

Thanks,
Wayne




From aj10 at my-deja.com  Tue Nov 30 18:36:02 1999
From: aj10 at my-deja.com (aj10 at my-deja.com)
Date: Tue, 30 Nov 1999 23:36:02 GMT
Subject: Removing null items
References: <821761$s8u$1@nnrp1.deja.com>
Message-ID: <821n12$8in$1@nnrp1.deja.com>

In article <821761$s8u$1 at nnrp1.deja.com>,
  sragsdale at my-deja.com wrote:
> So I've got a very simple desire: to remove all false (I.E. None,
0, '',
> []) items from a list.
>
> As it is I've got this function which works fine, but it offends my
> sight.  Is there any better (or at least different) way to do this?
>
> #################################################################
> # return non-null items from the first list
> def foo(lst):
>     rlist = []
>     for i in lst:
>         if i:
>             rlist.append(i)
>     return rlist
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
>
This has been answered just recently. Repeating it in the right place
with an example.
--
Python 1.5.2 (#0, Apr 13 1999, 10:51:12) [MSC 32 bit (Intel)] on win32
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> a=[None, 'aj', 'mk', 10, {1:2, 'a':10}, {}, None, []]
>>> a
[None, 'aj', 'mk', 10, {1: 2, 'a': 10}, {}, None, []]
>>> filter(None, a)
['aj', 'mk', 10, {1: 2, 'a': 10}]
>>>
--
Thx
-aj


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



From hinsen at cnrs-orleans.fr  Tue Nov  2 05:05:06 1999
From: hinsen at cnrs-orleans.fr (Konrad Hinsen)
Date: 02 Nov 1999 11:05:06 +0100
Subject: vector?
References: <9910291655.ZM20772@eel.ucsd.edu>
Message-ID: <m3wvs1qkdp.fsf@chinon.cnrs-orleans.fr>

"Mona Wong" <mona at eel.ucsd.edu> writes:

> 	Anyway, to make vectors using python, is there a module or should I
> just make it out of arrays?

It depends on what you mean by "vector". If you mean a collection of
numbers with associated arithmetic etc., these are just 1-dimensional
arrays as implemented in Numerical Python. If you mean vectors as used
in geometrical calculations, you can find an implementation in my
ScientificPython package
(http://starship.python.net/crew/hinsen/scientific.html), along with
other classes that are useful for geometry (tensors and linear
transforms).
-- 
-------------------------------------------------------------------------------
Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.55.69
Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais
-------------------------------------------------------------------------------



From timo at bbn.com  Thu Nov 11 17:47:04 1999
From: timo at bbn.com (Tim O'Malley)
Date: Thu, 11 Nov 1999 17:47:04 -0500 (EST)
Subject: Production Web Deployment?
In-Reply-To: <3826C69F.B4E26B89@digicool.com>
References: <NBBBKMMANMGEEBHPJJECGEGIFJAA.dirkl@home.com>
	<3826C69F.B4E26B89@digicool.com>
Message-ID: <14379.18122.156471.644349@ghost.bbn.com>

hola.

Paul Everitt writes:
 > Hopefully over the next month we'll migrate Zope.org away from PCGI (a
 > Zope system for getting requests from the web server to the long-running
 > process) to FastCGI.  This will eliminate execution of small C "wrapper"
 > executable.

I probably shouldn't mention this, but ...

  Is Zope.org using Apache?
  There is a prototype (alpha) Apache module which embeds the
  functionality of PCGI into Apache.  If this worked for you, then you
  would maintain compatibility with PCGI.

just a thought..

-- 
.....................................................................
.   happy daze            .   Age and treachery will beat           .
.        -tim O           .       youth and skill every time.       .
.....................................................................



From mikael at isy.liu.se  Thu Nov 11 06:52:43 1999
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Thu, 11 Nov 1999 13:52:43 +0200 (MET)
Subject: Intelligent strptime?
In-Reply-To: <19991110171256.A5196@optiplex.palga.uucp>
Message-ID: <XFMail.991111135243.mikael@isy.liu.se>

Yet another "but in Sweden we do"...

On 10-Nov-99 Gerrit Holl wrote:
 >  Does there exist an intelligent strptime? "26/11/99", for example, can
 >  _only_
 >  mean the 26th November of this year, and the chance that "14:11:12" means
 >  14 november 1912 is very small. Is there a module/function that guesses the
 >  format? There isn't doubt very often, only for 12/11/99, US is 11 december,
 >  Europe is 12 November (isn't it?).
 >  http://www.python.org/mailman/listinfo/python-list

Europe is not homogeneous. Swedes prefer most significant first, so we
would write 991126. Note no slashes, always 6 figures. Zeros are used
at obvious places if some numbers are less than 10, that is April the 
1st 1999 would be 990401.

Swedes accept 26/11/99, because it's still in order of significance,
but sighs "Americans" in the presence of 11/26/99, and we never know
what Americans mean by 1/4/11.

Someone pointed out that it would be bad if the program accepted one
format sometimes, and sometimes not. I couldn't agree more. I would 
probably throw my computer out the window, or if I took the time to
think a bit, only throw the software out the window. 

This problem is in the same division as points, commas, and blanks
in formatting numbers. The same old PITA.

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    11-Nov-99
Time:    13:33:39

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



From mlh at vier.idi.ntnu.no  Tue Nov 23 20:55:52 1999
From: mlh at vier.idi.ntnu.no (Magnus L. Hetland)
Date: 24 Nov 1999 02:55:52 +0100
Subject: help me learn please!!!!!
References: <81aj0m$fmg$1@ins23.netins.net>
Message-ID: <y0j7lj8irfb.fsf@vier.idi.ntnu.no>

"Charles Moore" <cmoore at netins.net> writes:

> I'd like to learn how to program are their any good suggestions as to were
> to start I'm a complete beginner I ask of you to point me in the right
> direction
> icq#54741375

Try

   http://www.idi.ntnu.no/~mlh/python/hacking.html

for a very basic introduction to programming in Python.

shameless-plug-but-at-least-it's-for-free'ly yours,

Magnus

:)

--

  Magnus          Echelon jamming noise:
  Lie             FBI CIA NSA Handgun Assault Bomb Drug Terrorism
  Hetland         Special Forces Delta Force AK47 Hillary Clinton 



From cgw at fnal.gov  Tue Nov  2 10:50:02 1999
From: cgw at fnal.gov (Charles G Waldman)
Date: Tue, 2 Nov 1999 16:50:02 +0100 (MET)
Subject: (no subject)
In-Reply-To: <381F00E1.721BB47F@cornell.edu>
References: <381F00E1.721BB47F@cornell.edu>
Message-ID: <14367.2090.901950.306647@gargle.gargle.HOWL>

Matthew Hirsch writes:
 > Hi All,
 > 
 > I'm trying to do something like the following:
 > 
 > a=4.0
 > b=2.0
 > print '2(%3.1f)' % a,b
 > 
 > where I really mean
 > 
 > print '%3.1f %3.1f' % a,b
 > 
 > but it's printing
 > 
 > 2(4.0) 2.0
 > 
 > Is there a way of doing this so I don't have to write out %3.1f,
 > multiple times?
 > 

Try this construction:

print 2*"%3.1f " % (a,b)


Note that "%" binds more tightly than "," so if you want to print
multiple vars in one format string, you must say (for example)
 
print "%s %s %s" % (a,b,c)

rather than:

print "%s %s %s" % a, b, c






From phd at phd.russ.ru  Mon Nov  1 04:11:15 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Mon, 1 Nov 1999 09:11:15 +0000 (GMT)
Subject: HTMLParser
In-Reply-To: <381C8DAB.511386A0@home.com>
Message-ID: <Pine.LNX.4.20.9911010908350.12533-100000@fep132.fep.ru>

On Sun, 31 Oct 1999, Edward Muller wrote:
> Does anyone have any sample code for the htmllib/HTMLParser library?

   http://sun.med.ru/~phd/Software/Python/
   http://members.xoom.com/phd2/Software/Python/
   http://www.fortunecity.com/skyscraper/unix/797/Software/Python/

   Download bookmarks_db.tgz. It is a set of programs to manipulate
bookmarks.html. One program, bkmk2db, splits bookmarks.html into a database
or text file, using HTMLParser.

Oleg.
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From fredp at mygale.org.nospam  Fri Nov 19 09:46:19 1999
From: fredp at mygale.org.nospam (Fred Pacquier)
Date: 19 Nov 1999 14:46:19 GMT
Subject: Bobo tarball, anyone ?...
References: <80v5g8$5a6$1@mulga.cs.mu.OZ.AU> <3834101A.FF39F64C@infercor.no>
Message-ID: <8E83A8056PaCmAnRDLM@194.2.0.33>

Speaking of Bobo... its pages at digicool.com are gone.
Can anyone point me to a place where I could get the last Bobo archive ?
(I need to run some ready-made scripts that depend on pcgi and bobo)
Thanks !

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



From neelk at brick.cswv.com  Thu Nov  4 20:22:23 1999
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 5 Nov 1999 01:22:23 GMT
Subject: Some more Ruby observations (Re: Comparison between Python and "Ruby")
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp> <7vjtrc$hmm$1@newshost.accu.uu.nl>  <3821A860.5D32AFCC@compaq.com>
Message-ID: <slrn824eo2.lsi.neelk@brick.cswv.com>

Greg Ewing <greg.ewing at compaq.com> wrote:
>Here are some random observations from studying the Ruby
>source code:
>
>* It doesn't use bytecode. The evaluator seems to operate
>  directly on a parse tree. I don't know what implications
>  this has for execution speed.

I was under the impression that Ruby did tail-recursion elimination,
and since continuation-passing forms are just a hair away from
bytecode, I am a bit surprised by this news. I'm not saying this is
impossible (witness the scm Scheme interpreter), mind, just a little
surprised.

Are there any syntax-tree munging phases to do stuff like common
subexpression elimination and loop hoisting? I guess that could yield
enough speedups to make a VM not yet necessary.


Neel



From news at dorb.com  Tue Nov 23 14:19:48 1999
From: news at dorb.com (Darrell)
Date: Tue, 23 Nov 1999 14:19:48 -0500
Subject: Visitor design pattern for python objects?
References: <818512$ma9$1@nnrp1.deja.com>
Message-ID: <ELB_3.9327$JD1.663967@typhoon.nyroc.rr.com>

How useful is the visitor pattern in Python ?
Doesn't this solve problems associated with strongly typed languages ?

One useful feature might be had if instead of visitor.visit(self)
you use visitor.myConcreteType(self)

That gives you a switch on type thing so you can avoid this:
    if type(concreteType)==type(xxxx):
        pass
    elif type(......


#!/usr/bin/env python
"""
My twisted view of visitor in Python
Darrell Gallion
"""

class Visitor:
    def __init__(self):
        self._methodDic={}

    def default(self, other):
        print "What's this:", other

    def addMethod(self, method):
        self._methodDic[method.getKey()]=method

    def __call__(self, other):
        method=self._methodDic.get(\
                other.__class__.__name__,self.default)
        return method(other)

class MyVisit:
    """
    Instead of deriving from Visitor the work is
    done by instances with this interface.
    """
    def __init__(self, otherClass):
        self._msg='Visit: %s'%otherClass.__name__
        self._key=otherClass.__name__

    def __call__(self, target):
        print self._msg, target

    def getKey(self):
        return self._key

# Some stuff to visit
class E1:pass
class E2:pass
class E3:pass

collection=[E1(), E1(), E2(), E3()]

visitor=Visitor()
visitor.addMethod(MyVisit(E1))
visitor.addMethod(MyVisit(E2))

map(visitor, collection)

###########################    OUTPUT:

Visit: E1 <__main__.E1 instance at 7ff6d0>
Visit: E1 <__main__.E1 instance at 7ff730>
Visit: E2 <__main__.E2 instance at 7ff780>
What's this: <__main__.E3 instance at 7ff7b0>

--
--Darrell





From adamv at nwlink.com  Sun Nov 14 21:34:26 1999
From: adamv at nwlink.com (Adam Vandenberg)
Date: Sun, 14 Nov 1999 18:34:26 -0800
Subject: Variable scope question.. singletons?
Message-ID: <382f70a7@news.nwlink.com>

I have a bunch of modules that will be used by a bunch of top-level scripts
(cgis).
Each top-level script is going to: "import prefs" which I want to check
cookies
and fill out a Prefs class.
So far OK.

Now, is there any way to make this class "more global" so that any of the
modules I import can access it?
I guess this is sort of like declaring the variable "extern".

Any thoughts? Does this make sense?

Thanks all.






From quinn at pfennig.ugcs.caltech.edu  Wed Nov 17 19:51:29 1999
From: quinn at pfennig.ugcs.caltech.edu (Quinn Dunkan)
Date: 18 Nov 1999 00:51:29 GMT
Subject: Alphabetized dictionary listings
References: <3832b6ce.80769668@news1.mysolution.com>
Message-ID: <slrn836je0.2f9.quinn@pfennig.ugcs.caltech.edu>

On Wed, 17 Nov 1999 14:23:51 GMT, Ken Power <iam at not.you> wrote:
>Why does this occur? More to the point, how can I have the print-out
>be in alphabetical order?
>
>>>> myList = ['Abel', 'Dexter', 'Francis', 'Matthew', 'Samantha', 'Tyrone']
>>>> myDict = {}
>>>> for name in myList:
>	myDict[name] = 0
>	
>	
>>>> for key in myDict.keys():
>	print key
>	
>	
>Tyrone
>Samantha
>Francis
>Abel
>Matthew
>Dexter

Both of your examples assume there is some sort of order to a dictionary...
there isn't.  The documentation for the keys() method in the reference manual
at 2.1.6 says: "Keys and values are listed in random order."  This has to do
with the way dictionaries are implemented internally.

sort() to the rescue!  The usual idiom for what you want is:

k = dict.keys()
k.sort()
for i in k:
    print i

If you're wondering why you can't just do

for i in dict.keys().sort():
    print i

check out faq 6.20

python-doesn't-allow-!s-in-identifiers-ly y'rs



From calishar at *NOSPAM*home.com  Tue Nov 30 18:29:02 1999
From: calishar at *NOSPAM*home.com (Calishar)
Date: Tue, 30 Nov 1999 23:29:02 GMT
Subject: WinNT and DLL's
References: <ddd64s4pp498hb7qlrgnq6o8fdqilrr27d@4ax.com> <001701bf3b3b$e7a39570$3acbd9c2@peridot.optichrome.com>
Message-ID: <2dn84ssgdboo47m0dqaavii70badgl6ctj@4ax.com>

On Tue, 30 Nov 1999 14:05:06 -0000, "Adrian Eyre"
<a.eyre at optichrome.com> wrote:

>>   Is there a way that I can simply access the pre-compiled dll files?
>
>Have a look at the calldll module: http://www.nightmare.com/software.html

 Thanks, got the same tip in my mailbox first thing this morning (my
brain sort of clued in to what it was, but my eyes were almost
screaming 'LOOK...'

Played with it a bit during the day, looks like it (and struct) will
do what I need.

Thanks for all who came up with this idea.

Calishar



From mlh at vier.idi.ntnu.no  Mon Nov 15 18:11:14 1999
From: mlh at vier.idi.ntnu.no (Magnus L. Hetland)
Date: 16 Nov 1999 00:11:14 +0100
Subject: Python module to format text into columns?
References: <80oq13$ql3$1@nnrp1.deja.com>
Message-ID: <y0jogcvic4t.fsf@vier.idi.ntnu.no>

hamish_lawson at my-deja.com writes:

> Is there a Python module that will take a list of
> strings and wordwrap them columnwise, similar to
> Perl's format mechanism? For example, if supplied
> with the following strings
> 
>    "14/11/1999"
>    "Here is a a paragraph of text."
>    "Here is another paragraph of text."
> 
> and with appropriate specification of column
> widths and alignments, it could produce something
> like:
> 
>    14/11/1999   Here is a   Here is another
>                 paragraph   paragraph of text.
>                 of text.

Without really knowing whether there is such a module, I offer a
simple implementation of some of the functionality, which can easily
be extended:

---- format.py ----



left   = 0
center = 1
right  = 2

def align(line,width=70,alignment=left):
    space = width - len(line)
    if alignment == center:
        return ' '*(space/2) + line + \
               ' '*(space/2 + space%2)
    elif alignment == right:
        return ' '*space + line
    else:
        return line + ' '*space


def split(paragraph,width=70,alignment=left):
    result = []
    import string
    words = string.split(paragraph)
    current, words = words[0], words[1:]
    for word in words:
        increment = 1 + len(word)
        if len(current) + increment > width:
            result.append(align(current,width,alignment))
            current = word
        else:
            current = current+" "+word
    result.append(align(current,width,alignment))
    return result


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

What remains to be done, is the parsing of a format string (or
"picture") and - if you like, to add the handling of multiple
paragraph (separated by a double newline) and optionally "squeezing"
several empty lines into one etc. Parsing the format string might take
a bit of work, but the rest should be trivial.

> 
> 
> Hamish Lawson

--

  Magnus          Echelon jamming noise:
  Lie             FBI CIA NSA Handgun Assault Bomb Drug Terrorism
  Hetland         Special Forces Delta Force AK47 Hillary Clinton 



From mcfletch at vrtelecom.com  Tue Nov 16 15:53:56 1999
From: mcfletch at vrtelecom.com (Mike Fletcher)
Date: Tue, 16 Nov 1999 15:53:56 -0500
Subject: list.without()?
In-Reply-To: <3831BE57.366ED454@appliedbiometrics.com>
Message-ID: <NDBBIAPHEKGOBIHCBLIIIEOOCAAA.mcfletch@vrtelecom.com>

Shouldn't make a big difference, as they exception only occurs once (when
you're out of elements to remove).  Doesn't challenge the front-runners
anyway.

Enjoy,
Mike

# Following just ran the original without and Chris' as without6

p:\>without
Testing function: <function without at 7d7f00>
  single element at middle:
    10: 0.0
    100: 0.0
    1000: 0.0
    10000: 0.00999999046326
    100000: 0.0699999332428
    1000000: 1.11199998856

  all elements being removed:
    10: 0.120000004768
    100: 0.0
    1000: 0.0100001096725
    10000: 0.770999908447
    100000: 78.7929999828
###Note: Tests skipped due to time overrun

Testing function: <function without6 at 7e0340>
  single element at middle:
    10: 0.0
    100: 0.0
    1000: 0.0
    10000: 0.0
    100000: 0.0710000991821
    1000000: 1.10099995136

  all elements being removed:
    10: 0.129999995232
    100: 0.0
    1000: 0.00999999046326
    10000: 0.761000037193
    100000: 79.8550000191
###Note: Tests skipped due to time overrun

-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Christian Tismer
Sent: November 16, 1999 3:28 PM
To: Mike Fletcher
Cc: python-list at python.org
Subject: Re: list.without()?




Mike Fletcher wrote:
>
> Hmm, actually, it's only cheap for certain subsets of the problem (which I
> took to be the common case).  In particular, it shines for small numbers
of
> instances in lists of fairly arbitrary size.  It falls down completely
when
> you are removing large numbers of instances (such as all of them).  I just
> tested five algos, testing 1 instance in middle of list, and a list
composed
> of only elements to be removed, findings, module, and raw test results
> below...

Oh well, I overlooked that you have the try inside the while.
That's bad. I thought I had read this:

def without( source, element):
        temp = source[:]
        try:
                while temp:
                        temp.remove( element )
        except:
            return temp
        return temp

Did you try this (I didn't, have no time).

cheers - chris

--
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home

--
http://www.python.org/mailman/listinfo/python-list




From skip at mojam.com  Sun Nov 21 15:09:24 1999
From: skip at mojam.com (Skip Montanaro)
Date: Sun, 21 Nov 1999 14:09:24 -0600 (CST)
Subject: Order of keyword arguments
In-Reply-To: <38380F92.D9817451@corrada.com>
References: <38379233.ADE65CDB@corrada.com>
	<19991121034245.A21255@quark.emich.edu>
	<38380F92.D9817451@corrada.com>
Message-ID: <14392.20852.400630.148814@dolphin.mojam.com>

    Andres> I want to have my cake and eat it too! I thought this would be a
    Andres> way of keeping the flexibity of keyword arguments but also pass
    Andres> information along on how the calling function wants the items
    Andres> returned. For example, an HTML form could be called with a
    Andres> variable number of input tags and you want to render the HTML in
    Andres> the same order that they were used to call the function.

I solved this in one instance by giving all the input tag names names like
XXXtagname, where XXX is a zero-padded number.  The order of display is
determined by the form's author, so you could have

    001name
    002address
    003address
    004city
    005state

then in the cgi handler you simply sort them, then strip the first three
characters. 

Another thing to check if you have access to it is the raw QUERY_STRING.  I
believe browsers are supposed to (or at least do by convention) assemble the
string based upon the order of arguments in the form.  That way you can have
two fields named "address" and know that the order you see will be the same
as the order in which the user typed their street address and suite number.

Skip Montanaro | http://www.mojam.com/
skip at mojam.com | http://www.musi-cal.com/
847-971-7098   | Python: Programming the way Guido indented...




From bernhard at alpha1.csd.uwm.edu  Tue Nov 16 15:29:46 1999
From: bernhard at alpha1.csd.uwm.edu (Bernhard Reiter)
Date: 16 Nov 1999 20:29:46 GMT
Subject: wxPython on Solaris??
References: <942778029.608457@jerusalem.magic.fr>
Message-ID: <slrn833flq.6jm.bernhard@alpha1.csd.uwm.edu>

On Tue, 16 Nov 1999 19:42:46 +0100, Fran?ois Desvall?es <francois at roya.org> wrote:

>I need to to display graphics (raw bitmaps) on a Solaris workstation. I am
>thinking of using wxPython, but I have only seen reports of it being used on
>Windows and Linux. Has this been done, and how much work would it require
>(using the Motif interface, I guess?)

It probably has been done. (Though I didn't try it myself.)
You would use wxGtk and the GTK lib. 
Gtk works on Solaris. (I saw it.)


I would start at the wxPython documentation. http://alldunn.com/wxPython/
	Bernhard
-- 
Research Assistant, Geog Dept UM-Milwaukee, USA.  (www.uwm.edu/~bernhard)
Free Software Projects and Consulting 		         (intevation.net)  
Association for a Free Informational Infrastructure            (ffii.org)



From achrist at easystreet.com  Wed Nov 24 16:20:59 1999
From: achrist at easystreet.com (Al Christians)
Date: Wed, 24 Nov 1999 13:20:59 -0800
Subject: More problems with random seed (was Re: whrandom seed question)
References: <81eu22$411al$1@umbc7.umbc.edu> <m366ysu7t4.fsf@atrus.jesus.cam.ac.uk> <81hej5$3tmv0$1@umbc7.umbc.edu>
Message-ID: <383C56BB.8B7C608F@easystreet.com>

Stelios Kyriacou wrote:
> 
> Why when I call the gauss distribution I get 2 alternating values:
> 
> >>> random.seed(0); print random.gauss(0,1);
> 2.11224018067
> >>> random.seed(0); print random.gauss(0,1);
> 0.225551631579
> >>> random.seed(0); print random.gauss(0,1);
> 2.11224018067
> >>> random.seed(0); print random.gauss(0,1);
> 0.225551631579
> >>> random.seed(0); print random.gauss(0,1);
> 2.11224018067
> >>>
> 
> Any more ideas?


Surprisingly:

>>> random.seed(0)
>>> random.gauss(0,1) / random.gauss(0,1)

Gives me either 0.106783136522 or 9.36477455684

And, astoundingly (???):

>>> 9.36477455684 * 0.106783136522
1.0

Not really, according to Numerical Recipes, etc, random gaussian's are 
generated in pairs.  One is saved for next time.  The random.seed() 
function must not be clearing out the cached value.

Should it?

Al



From bferris at my-deja.com  Tue Nov 16 05:45:21 1999
From: bferris at my-deja.com (bferris at my-deja.com)
Date: Tue, 16 Nov 1999 10:45:21 GMT
Subject: Telnet server
Message-ID: <80rck0$n3i$1@nnrp1.deja.com>

Anyone out there seen a telnet SERVER written in python?  Sure, I could
write one but I'll only get the urge to do so if nobody has one on
offer.


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



From lalal at fd.com  Thu Nov 25 19:14:07 1999
From: lalal at fd.com (lalal at fd.com)
Date: Thu, 25 Nov 1999 18:14:07 -0600
Subject: make money for doing nothin! 47438
Message-ID: <251199181407@fd.com>

Everyone likes making easy money.  This place pays you 50 cents for every hour you browse the web!!!
Get even more if you refer people!  I just got a check for $36.00!!!

Sign up here:
http://www.alladvantage.com/go.asp?refid=GOB631








(E-4aLH=]X



From skip at mojam.com  Sat Nov 20 09:21:19 1999
From: skip at mojam.com (Skip Montanaro)
Date: Sat, 20 Nov 1999 08:21:19 -0600 (CST)
Subject: Naming: fp
In-Reply-To: <19991120130649.A18627@optiplex.palga.uucp>
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net>
	<3d1z9sq2uc.fsf@amarok.cnri.reston.va.us>
	<JlFX3.3659$s87.67930@dfiatx1-snr1.gtei.net>
	<slrn82unnr.mja.thor@localhost.localdomain>
	<19991120130649.A18627@optiplex.palga.uucp>
Message-ID: <14390.44639.812811.317191@dolphin.mojam.com>

    >> A file object fp that  points to the mailbox file.
    Gerrit> Why is a file object often called 'fp'?

Stdio's basic data structure is a struct called FILE.  It's always exposed
as a FILE * (or file pointer), hence the propensity to name variables of
type FILE * "fp":

     #include <stdio.h>
     ...
     char * filename = "/tmp/trash";
     FILE *fp = fopen(filename);
     ...

Skip Montanaro | http://www.mojam.com/
skip at mojam.com | http://www.musi-cal.com/
847-971-7098   | Python: Programming the way Guido indented...



From g_will at cyberus.ca  Mon Nov 15 22:31:47 1999
From: g_will at cyberus.ca (Gordon Williams)
Date: Tue, 16 Nov 1999 03:31:47 GMT
Subject: Line Benchmarks Tkinter vs. wxPython
References: <01bf2fa7$02992ca0$8a42c3d1@Cyberus.cyberus.ca> <slrn8318ec.9dg.bernhard@alpha1.csd.uwm.edu>
Message-ID: <01bf2fda$9bf27f40$9c55c3d1@Cyberus.cyberus.ca>

Hi,

Sorry, I should have given you the os as well.  I was running it under W95.

If you need some more info, drop me a line.  Maybe someone else using Unix
could make a similar comparision.

Regards,

Gordon

Bernhard Reiter <bernhard at alpha1.csd.uwm.edu> wrote in article
<slrn8318ec.9dg.bernhard at alpha1.csd.uwm.edu>...
> 
> On Mon, 15 Nov 1999 20:22:16 GMT, Gordon Williams <g_will at cyberus.ca>
wrote:
> >I have done some benchmark testing to see the difference in speed
between
> >plotting multi point lines using Tkinter and wxPython.  
> Very nice. I appreciate the effort.
> 
> I couldn't figure out on which window platform you tried this.
> 
> wxGtk or wxWindows?
> 
> as the wxWindows compatible libraries use native windows calls
> this might make a significant difference.
> 
> 	Bernhard
> -- 
> Research Assistant, Geog Dept UM-Milwaukee, USA.  (www.uwm.edu/~bernhard)
> Free Software Projects and Consulting 		         (intevation.net)  
> Association for a Free Informational Infrastructure            (ffii.org)
> 



From aahz at netcom.com  Tue Nov 16 11:51:36 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 16 Nov 1999 16:51:36 GMT
Subject: Beginning programming with Python
References: <sng0ychgwt.fsf@motorola.com> <80p9ds$ppo$1@newshost.accu.uu.nl> <asteroidal-80psfo/INN-2.2.1/bat@broadway.news.is-europe.net> <38317787.19AFDAF5@compaq.com>
Message-ID: <80s22o$o0l$1@nntp9.atl.mindspring.net>

In article <38317787.19AFDAF5 at compaq.com>,
Greg Ewing  <greg.ewing at compaq.com> wrote:
>Janos Blazi wrote:
>> 
>> the indenting this time? There is actually nothing else you can forget in
>> Python.
>
>Well, not quite - you can forget the "self" in method
>declarations fairly easily...

I don't seem to have that problem; my fingers have learned to
automatically type "(self" whenever "def" is indented.  ;-)
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From fredrik at pythonware.com  Fri Nov 12 10:28:51 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 12 Nov 1999 16:28:51 +0100
Subject: '==' vs. 'is' behavior
References: <382C2C6A.64A8F2C2@theriver.com>
Message-ID: <00b601bf2d22$a0cdc0c0$f29b12c2@secret.pythonware.com>

Andy Fox <foxes at theriver.com> wrote:
> I thought I understood the difference between the '==' and 'is'
> operators, but just found an exception to the rule.

nope.  no exception.  read on, to understand
why.

> I understand the following:
> >>> a = (1, 2)
> >>> b = (1, 2)
> >>> a == b #are they equivalent?  yes.
> 1
> >>> a is b #are they the same object? no.
> 0
 
> Why doesn't it work for a certain integers:
> >>> a = 2
> >>> b = 2
> >>> a == b #are they equivalent?  yes.
> 1
> >>> a is b #are they the same object? yes.  Why?
> 1

integers are immutable, which means that they
can be shared between variables.  python's integer
implementation happens to preallocate a set of
small integers, to speed things up a little.

but maybe you've missed that variables are just
names?

"a" and "b" happen to point to the same internal
integer object.  if you assign one of them to some-
thing else, you don't change the integer, you just
tell python to bind the name to something else.

in other words, it's perfectly possible to have
many variable names pointing to the same internal
object.

btw, the "id" function gives you a unique identity
for each object:

a = 2
b = 2
print id(a)
print id(b)

prints

8918500
8918500

(or something similar).

and "a is b" is of course the same thing as
"id(a) == id(b)",

> Further experimentation shows that this seems to work for integers
> through 99, but that starting with 100 the results of the 'is' test
> return 0, as I would expect.

python's integer implementation provide ready-made
integers for that range.  nothing strange here, really.

finally, sys.getrefcount allows you to check how many
others that happen to point to any one object:

import sys

for i in xrange(150):
    print i, sys.getrefcount(i)

prints:

0 55
1 40
2 22
3 8
4 14
5 4
6 4
7 5
8 10
9 4
10 3
11 5
12 3
13 3
14 3
15 5
94 3
...
95 3
96 3
97 3
98 3
99 3
100 2
101 2
102 2
103 2
104 2
105 2
...

on my box...  (for "100", there's one reference
from the "i" variable, and one extra introduced
by the getrefcount function itself -- it needs
to hold on to the integer object while deter-
mining the reference count...  if the count is
higher than 2, someone else is also referring
to the same value)

</F>




From ivanlan at callware.com  Tue Nov 16 16:28:01 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Tue, 16 Nov 1999 14:28:01 -0700
Subject: XpmImagePlugin.py in PIL
Message-ID: <3831CC61.A71E796D@callware.com>

Hi All--
I've been trying to fix the XpmImageFile class in XpmImagePlugin.py. 
I've got the colors working correctly (but not the color names, yet),
but I do not understand what this line does:

		self.tile = [("raw", (0, 0)+self.size, self.fp.tell(), ("P", 0, 1))]

And why the load_read() method is never called.  Load_read() seems (from
reading it) to have the correct way of reading the data for the image
in, but it's not called.  The self.tile line, on the other hand, seems
to be arguments for some function that sucks up the whole file as one
big hunk.  The skewed image I'm seeing is consistent with that
behaviour.

Any hints?  Where should I RTFM?

<if-there-are-one-i'd-LOVE-to-read-it!>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours
----------------------------------------------



From nick at src.uchicago.edu  Fri Nov  5 17:20:38 1999
From: nick at src.uchicago.edu (Nick Collier)
Date: Fri, 05 Nov 1999 16:20:38 -0600
Subject: Numbers in SQL statements
Message-ID: <38235836.D2F9A909@src.uchicago.edu>

Hi,

I've been creating SQL statements from dictionaries holding the field
name and the new value to insert or update. As might be expected this
new value is often a number. When turning this number into a string in
order to create the sql statement the number will have a "L" appended if
it is a long. Consequently, the resulting SQL might look like the
following:

update myTable
set id_num = 23L
where food = 'spam'

I've been testing if the value is a number whose string representation
ends with "L" and then stripping the "L", but am pretty sure there must
be a better way. If so, any suggestions as to what is?

thanks,

Nick
-- 

Nick Collier
nick at src.uchicago.edu
Social Science Research Computing
University of Chicago
Chicago, IL 60637



From rparker at bserv.com  Fri Nov  5 23:54:26 1999
From: rparker at bserv.com (Robert Parker)
Date: Sat, 06 Nov 1999 04:54:26 GMT
Subject: Printing on a Windows workstation
Message-ID: <6wOU3.142$dF2.175@198.235.216.4>

Hi All
    I have waded through the man pages and the functions to try and
determine what to use to print to a printer using Win32.  I don't know if
there is a COM that could be used or some other method to access the printer
API in Win32.  I don't feel excited about having to generate lines of escape
sequences to accomplish formatting and proper presentation.  Does any one
have any suggestions?

    As a newbie it sure is helpful having this newsgroup.  Am greatful for
any assistance.

Robert Parker-)





From x at x.x  Sat Nov  6 12:25:35 1999
From: x at x.x (Eric Jacobs)
Date: Sat, 06 Nov 1999 12:25:35 -0500
Subject: A newbie dictionary question
References: <byr9i42us9.fsf@bolzano.math.ethz.ch>
Message-ID: <3824648F.7FB08EAC@x.x>

Jan Vroonhof wrote:
> 
> I have the distinct feeling this is FAQ, but if so I missed it. It is
> late (or better early) here.
> I am trying to teach  myself python with a little problem I have that
> involves parsing ChangeLog files.
> 
> Thus I wrote a module that creates a dictionary-like class with the
> text of the ChangeLog entry as the key. Later I want to use this to
> check whether specific changes were applied. However I want to do the
> comparisons regardless of whitespace. Thus I a made small class that
> has __cpm__ and __hash__ functions that ignore whitespace.
> 
> However looking up objects in the hashes doesn't seem to work. Can
> somebody tell me what obvious thing I am missing?
> 
> The code is below. Note that I added the complete module (which is 20
> lines or so more than a trivial example) in the hope that somebody
> comment on it stylewise as this is my first Python program.
> In particular I have the feeling the while loops in the scan(self)
> function could be done much clearer
> 
> #!/usr/bin/env python
> """This class implements a dictonary like class,
> ChangeLog, that represents entries from ChangeLog file"""
> import re
> import sys
> import string
> from UserDict import UserDict
> 
> authpat = re.compile(
>     r"^(\d\d\d\d-\d\d-\d\d|\w\w\w \w\w\w\s+\d+\s+\d\d:\d\d:\d\d \d\d\d\d)"
>     +r"\s+([^<]+)\s+<(.+)>\s?$")
> emptypat = re.compile(r"^\s*$")
> startpat = re.compile("^(?:\t| {4})\\s*\\* (.*)")
> contpat =  re.compile("^\\s*(?:\t| {4})\\s*(\\S.+)")
> 
> _debug = 0
> 
> class ChangeAuthor:
>     def __init__(self,name,address):
>         self.name = name
>         self.address = address
> 
>     def __str__(self):
>         return self.name + "<" + self.address + ">"
> 
> class ChangeEntry:
>     def __init__(self,value):
>         self.value = value
> 
>     def __str__(self):
>         self.value

Unrelated, but did you mean "return self.value"? It's
legal to throw away values in Python.

> 
>     # bad approximation?
>     def nospaces(self):
>         return string.join(string.split(self.value))
> 
>     def __cmp__(self,other):
>         return self.nospaces() == other.nospaces()

The __cmp__ is returning the wrong thing. It should return
-1 for <, 0 for ==, or 1 for >. You probably want

      def __cmp__(self, other):
           return cmp(self.nospaaces(), other.nospaces())

The way it's written above, objects that are supposed to
be equal will appear one greater than the other, and this
will break it. Dictionaries require their key objects to
have consistent __cmp__ and __hash__ functions.



From stuart.hungerford at webone.com.au  Mon Nov  1 05:50:47 1999
From: stuart.hungerford at webone.com.au (Stuart Hungerford)
Date: Mon, 01 Nov 1999 21:50:47 +1100
Subject: How can I make my assertions smarter?
Message-ID: <381D7087.BF8510E5@webone.com.au>

Hi all,

I'm using some simple functions to implement crude precondition,
postcondition and invariant checking in Python classes.

(Tim: please be gentle ;-)

Is there some neat way to wrap these tests up so I can print any
code expression that fails, e.g:

     Precondition failed foo.py line 42: spoon.shape == 'bent'

which would appear at line 42 of foo.py as

     pre(spoon.shape == 'bent')

I can see how to use environment variables or global variables to
selectively enable precondtions etc, although any hints for doing that
would be appreciated too.

Thanks in advance,

Stu




From ionel at psy.uva.nl  Mon Nov 29 06:29:17 1999
From: ionel at psy.uva.nl (Ionel Simionescu)
Date: Mon, 29 Nov 1999 12:29:17 +0100
Subject: Recursion?
References: <38421570.9C3D9AEC@videotron.ca>
Message-ID: <81tnv4$o57@mail.psy.uva.nl>

Pierre-Yves Cloutier <pycloutier at videotron.ca> wrote in message
news:38421570.9C3D9AEC at videotron.ca...
| Can I use recursion in python?

Yes.

---

def fact(n):
    if n==1: return 1
    else: return n*fact(n-1)

print fact(10)

---

ionel





From frcontin at tin.it  Tue Nov  2 18:43:58 1999
From: frcontin at tin.it (Francesco)
Date: Wed, 3 Nov 1999 00:43:58 +0100
Subject: Please Help!!!
Message-ID: <7vnsjg$jih$1@nslave1.tin.it>

I'm learning Python...
I'm building a simple telnet client...
I don't know how to catch the key pressed without read it from stdin...
I read the manual about the Event class but i can't understand a lot...
Can you help me, maybe with an example?






From darcy at vex.net  Thu Nov 11 13:14:10 1999
From: darcy at vex.net (D'Arcy J.M. Cain)
Date: 11 Nov 1999 18:14:10 GMT
Subject: ASCII delimited files
References: <942231502.1914033961@news.interbusiness.it> <80ekhb$1qrh$1@hub.org> <80et0b$2df8$1@hub.org> <80evr8$2lc7$1@hub.org>
Message-ID: <80f11i$2nof$1@hub.org>

D'Arcy J.M. Cain <darcy at vex.net> wrote:
> Skip it.  There's a bug.  If a line ends with a quoted entry it gets
> attached to the next one.  I'm tracking it down now and will post the
> corrected one.

OK, this one does the right thing.

#! /usr/bin/env python
# delimited.py
# Modified from the original by D'Arcy J.M. Cain <darcy at druid.net>
#
# CT990226 V0.1

"""
breaking tab delimited or CSV text into a list of tuples.
Original version: SMC databook plugin, readxls.py
Generalized to be usable for any delimiter but \n.
"""

import string, re

def split_delimited(s, delimiter=None) :
	"""split a delimited text file string into a list of tuples.
	Quotes are obeyed, enclosed newlines are expanded to tab,
	double quotes go to quotes"""

	# 980426 finding line delimiter dynamically

	probe = s[:10000]
	eol = findlinedelimiter(probe)

	# 990226 guessing field delimiter from '\t,' if not supplied

	if not delimiter:
		candidates = [
			(string.count(probe, '\t'), '\t'),
			(string.count(probe, ','), ','),
		]
		candidates.sort()
		delimiter = candidates[-1][-1]
	del probe

	# the trick to make this handy is to use SEP as a placeholder

	# Kind of ugly but it works for embedded commas - DJMC
	inquote = 0
	SEP='|'
	for i in range(len(s)):
		if s[i] == '"': inquote = (inquote == 0)
		if inquote == 0 and s[i] == delimiter:
			s = s[:i] + SEP + s[i + 1:]
	parts = string.split(s, '"')
	limits = (0, len(parts)-1)
	for i in range(len(parts)) :
		part = parts[i]
		if i%2 :
			part = string.replace(part, eol, delimiter)
		else :
			if not part and i not in limits: part = '"'
		parts[i] = part
	# merge it back
	txt = string.join(parts, "")
	parts = string.split(txt, eol)
	# now break by SEP
	for i in range(len(parts)) :
		# strip leading and trailing spaces
		#fields = string.split(parts[i], SEP)
		fields = re.split(r' *\0 *', parts[i])
		parts[i] = tuple(fields)
	return parts

# utilities

def findlinedelimiter(txt) :
	"""
	provide some kb of text to this function. It will determine
	the best delimiter and therefore guess the system
	"""
	mac  = "\x0D"
	unix = "\x0A"
	dos  = mac+unix
	oses = [dos, unix, mac]
	# find the one which gives the most lines.
	# in doubt, the longest delimiter wins.
	lis = []
	while txt and txt[-1] in dos: txt = txt[:-1] # CT970904
	for delim in oses:
		lis.append((len(string.split(txt, delim)), delim))
	lis.sort();
	return lis[-1][-1]

if __name__=="__main__":
	for l in  split_delimited('''1, 2,3, "vier", "quo""te", "embedded, comma", "this
is with a newline", here
another record
And another, "should end here"
"should not be attached to previous"'''):
		print l

# eof



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



From claudius at catlover.com  Tue Nov  9 23:38:45 1999
From: claudius at catlover.com (claudius at catlover.com)
Date: Wed, 10 Nov 1999 04:38:45 GMT
Subject: Java vs. Python - Serialization Bakeoff
References: <99110918085909.01567@quadra.teleo.net> <mD6W3.25292$Ec1.280622@news1.frmt1.sfba.home.com>
Message-ID: <pF6W3.25297$Ec1.279299@news1.frmt1.sfba.home.com>

claudius at catlover.com says:
>Patrick Phalen <python-list at teleo.net> says:
>>http://starship.python.net/crew/gandalf/DNET/serial/jp-serial.html
>>contains an report on tests comparing Java's and Python's data
>>marshalling, using various criteria. Numerous interesting charts and
>>graphs are presented.
>>
>>Since this was published in 1996 (Python 1.4), I wonder if anyone
>>expert in this subject has any comment on whether the intervening years
>>would have changed the conclusions substantively -- namely that Python
>>is, on average, faster, simpler, and more flexible.
>
>Hmm?  I get 'not found' (and I'm very interested in reading your findings.)

Ah, I found it...for those interested:
http://starship.python.net/crew/gandalf/DNET/serial/jp-serial.htm

(drop the trailing 'l')



From mwh21 at cam.ac.uk  Fri Nov  5 10:51:00 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 05 Nov 1999 15:51:00 +0000
Subject: return boolean from functions
References: <7vrq9b$hmr$1@nnrp1.deja.com> <7vsvef$elo$1@nnrp1.deja.com> <7vuoqt$m79$1@nnrp1.deja.com>
Message-ID: <m3bt999btn.fsf@atrus.jesus.cam.ac.uk>

Preston Landers <prestonlanders at my-deja.com> writes:

> In article <7vsvef$elo$1 at nnrp1.deja.com>,
>   Preston Landers <prestonlanders at my-deja.com> wrote:
> 
> > Alternatively, you can do this:
> >
> > TRUE=1
> > FALSE=0
> >
> > def foo(bar):
> >   if bar == "spam":
> >      return TRUE
> >   else:
> >      return FALSE
> 
> Sorry to follow up my own article, but it should be noted that the way I
> have it above will cause a namespace lookup (TRUE and FALSE) every time
> you call foo() (aside from the usual function overhead.)  So, if you are
> calling this function from a critical loop, you might want to avoid the
> namespace lookup and just return None or 1.  Also note that the word
> "None" is optional when you want to return false; you can do this:
> 
>   if foo:
>      return  # will return None

Umm, `return None' causes a name lookup too, and one that will wind up
getting resolved in __builtins__ - the searching place of last resort,
so

def f1():
    return None

will be slower than

FALSE = 0
def f2():
    return FALSE

which will be slower than

def f3():
    return 0

which will be about the same as

def f4():
    return

which wont be much different to

def f5():
    pass

which in turn will be ever-so-slightly slower than

def f6(None=None):
    return None

(which I bet you didn't expect!)

It's worth noting (unlike the above) that the differences involved are
very very small, so you should go for the one which expresses what you
actually mean, which imho is most likely to be f1 or f3.

there's-more-than-one-way-to-say-nothing-ly y'rs
Michael



From sekter at mail.matav.hu  Thu Nov 25 11:47:02 1999
From: sekter at mail.matav.hu (Arpad Kiss)
Date: Thu, 25 Nov 1999 17:47:02 +0100
Subject: Listing a dictionary
References: <383CFEE5.951376A0@obop.com.pl>
Message-ID: <013c01bf3764$baee2610$029f38c3@geometria.hu>

----- Original Message -----
From: Przemys?aw G. Gawro?ski <gawron at obop.com.pl>
To: <python-list at python.org>
Sent: Thursday, November 25, 1999 10:18 AM
Subject: Listing a dictionary


> How can I list all the keys in a dictionary ?
>
Dictionary has a keys method.
(http://www.python.org/doc/current/lib/typesmapping.html )


Arpad




From m.faassen at vet.uu.nl  Thu Nov 25 22:21:54 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 26 Nov 1999 03:21:54 GMT
Subject: Python complaints
References: <000401bf3733$af411100$be2d153f@tim>
Message-ID: <81kuci$33g$2@newshost.accu.uu.nl>

Tim Peters <tim_one at email.msn.com> wrote:

[if foo: bar() is ugly]
> [Martijn Faassen]
>> Agreed. The Zope sources are full of the former and that idiom
>> doesn't help me read them particularly.
>>
>> Perhaps 1.6 could introduce this requirement, and some scripts to
>> convert existing code. The only use I can see for the former is to
>> write nice obfuscated one liners. :)

> There are a few things in the language primarily for convenience in
> interactive mode.  This is one of them.

Aah, I see. Hm. Should be some way to allow something like this in 
interactive mode and not in source text. Probably a tricky problem, however.

> Being Dutch, Guido grossly
> overestimated the common sense of his American audience <wink>.

Ah, I wonder if that's what I've been doing too! :) Anyway, I have accesa
to the Zope sources; perhaps they'll accept my patches.

not-that-I-have-any-audience-ly yours,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?



From fdrake at acm.org  Thu Nov 11 12:10:46 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Thu, 11 Nov 1999 12:10:46 -0500 (EST)
Subject: .html -> .ps ?
In-Reply-To: <19991111120757.C1783@better.net>
References: <19991111080506.A1007@better.net>
	<14378.54900.714487.334690@weyr.cnri.reston.va.us>
	<19991111120757.C1783@better.net>
Message-ID: <14378.63638.48886.564135@weyr.cnri.reston.va.us>

William Park writes:
 > Found the solution by doing 'netscape -help'... duh!  See
 >     http://home.netscape.com/newsref/std/x-remote.html
 > Basically,
 >     $ netscape -remote "openURL(http://...)"
 >     $ netscape -remote "saveAs(test.ps,Postscript)"

  Ah, so you don't mind having a window opened on the screen.  I
figured you were trying to get around that as well.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From boud at rempt.xs4all.nl  Sun Nov 14 15:26:07 1999
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 14 Nov 1999 20:26:07 GMT
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us>
Message-ID: <80n5sv$7a$1@essle.valdyas.org>

Andrew M. Kuchling <akuchlin at mems-exchange.org> wrote:

<...>

> http://www.vex.net/~x/parnassus/ is *very* nice, and it might be the
> only solution required.  Apparently a good deal of work has been put
> into seeding the initial database, and that's important, because it
> makes the index immediately useful.  Yet I hadn't heard about it until
> now; is the site still beta?  Did I miss a c.l.py.announce article
> about it?  

You don't say! I just went there, spurred on by your posting to
enter my own efforts. Just when I had done so, I noticed that one
of them was already there - really good work! It seems to be very,
very nice, indeed.

-- 

Boudewijn Rempt  | http://www.xs4all.nl/~bsarempt



From aahz at netcom.com  Wed Nov 10 12:10:48 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 10 Nov 1999 17:10:48 GMT
Subject: Text Manipulation in Python
References: <VA.00000012.055a8702@ops.corpex.com> <19991110063134.A3633@better.net>
Message-ID: <80c8uo$r2$1@nntp9.atl.mindspring.net>

In article <19991110063134.A3633 at better.net>,
William Park  <parkw at better.net> wrote:
>
>But, if you're running Unix, you can also use 'sed'.

Even if you're not running Unix, you can also use sed.
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From r.albanese at qut.edu.au  Thu Nov 11 22:17:35 1999
From: r.albanese at qut.edu.au (Rico Albanese)
Date: Fri, 12 Nov 1999 13:17:35 +1000
Subject: Random text generator
Message-ID: <382B86CF.AA1C12C9@qut.edu.au>

Can anyone tell me if there is a random text generator for Python?  I
have an application that needs to generate a random password for users.
I have looked at the random and whrandom modules but they seem to create
random numbers only.

Thanks in advance for any advice given

Rico





From karanvasudevaNOkaSPAM at usa.net.invalid  Sat Nov  6 19:55:06 1999
From: karanvasudevaNOkaSPAM at usa.net.invalid (Karan Vasudeva)
Date: Sat, 06 Nov 1999 16:55:06 -0800
Subject: exec / execfile namespace problem
References: <004aa0e3.c3e96526@usw-ex0101-002.remarq.com> <m3904baoej.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <000b8d9b.b9d3bdca@usw-ex0101-003.remarq.com>

Thanks for responding. I believe I missed out on the real issue here.
You're right, doing a chunk++ like that would be inefficient, but
being new at this, I was kind of looking at getting the whole thing
done first. ok, what i meant was, i'm trying to exec (or execfile)
a script that  does something like this:

> > import httplib
> > h = httplib.HTTP('')
> > h.putrequest('GET', '/favicon.ico')
ok, you can fill in the host here, but that doesn't matter. my problem
is, when i call my script from another script (by doing an exec, thats
what i need to do, you see), the interpreter stops at httplib.py
as it can't find that socket called self.sock . This appears to be a
namespace issue. I just want to know how to do this right.

Both "python cool.py" and exec (and execfile) work absolutely ok
when called interactively. It just doesn't work from another script.

Regards,
Karan.

In article <m3904baoej.fsf at atrus.jesus.cam.ac.uk>, Michael Hudson
<mwh21 at cam.ac.uk> wrote:
> Karan Vasudeva <karanvasudevaNOkaSPAM at usa.net.invalid> writes:
> > Hi,
> >
> > I'm trying to do this:
> > execfile('cool.py')
> >
> > and this:
> >
> > import fileinput
> > chunk = ''
> > for line in fileinput.input('cool.py'):
> >    chunk = chunk + line
> > exec(chunk)
> How does that differ from
> exec open('cool.py').read()
> ? What you have will be really quite inefficient, because
> repeatedly
> executing
>     chunk = chunk + line
> is a quadratic algorithm ... but back to the point.
> >
> > where 'cool.py' is this: (in part)
> > import httplib
> > import time
> > import whrandom
> > h = httplib.HTTP('')
> > h.putrequest('GET', '/favicon.ico')
> Have you tried executing `python cool.py'? The obvious (to me)
> problem
> with the above is that you're not telling httplib about server! I
> think you want
> h = httplib.HTTP("server.domain.tld")
> HTH,
> Michael



* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!




From aahz at netcom.com  Fri Nov  5 18:40:57 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 5 Nov 1999 23:40:57 GMT
Subject: Numbers in SQL statements
References: <38235836.D2F9A909@src.uchicago.edu>
Message-ID: <7vvpu9$csr$1@nntp4.atl.mindspring.net>

In article <38235836.D2F9A909 at src.uchicago.edu>,
Nick Collier  <nick at src.uchicago.edu> wrote:
>
>I've been testing if the value is a number whose string representation
>ends with "L" and then stripping the "L", but am pretty sure there must
>be a better way. If so, any suggestions as to what is?

if type(foo) = types.LongType :
  return ( str(foo)[:-1] )
else :
  return ( str(foo) )
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From mhammond at skippinet.com.au  Tue Nov 16 21:37:27 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 17 Nov 1999 02:37:27 GMT
Subject: Getting existing instance of Excel
References: <3832016f.10696885@news.ne.mediaone.net>
Message-ID: <HxoY3.5692$Xo.30354@news-server.bigpond.net.au>

Joel Gould wrote in message <3832016f.10696885 at news.ne.mediaone.net>...
>Using win32com, you can get a new instance of excel with a command
>line:
>
>app=win32com.client.Dispatch('Excel.Application')
>
>But if Excel is already running, is there a way to get a pointer to
>the running instance of Excel?

That does return the existing instance.  If Excel is running, the existing
instance is used.  I just confirmed this using the task manager.

Indeed, I am not aware of how you can get COM to force a new instance of
Excel...

Mark.






From mcfletch at vrtelecom.com  Fri Nov 12 00:23:00 1999
From: mcfletch at vrtelecom.com (Mike Fletcher)
Date: Fri, 12 Nov 1999 00:23:00 -0500
Subject: ASCII delimited files
In-Reply-To: <034b01bf2c9a$ed7c3830$0100a8c0@rochester.rr.com>
Message-ID: <NDBBIAPHEKGOBIHCBLIICEOJCAAA.mcfletch@vrtelecom.com>

I never use CSV or TSV files myself (well, not more than once or twice per
year), I was just approaching the problem as one of optimisation (I'm an
optimisation junky).  The trailing empty record is actually there in the
dataset I'm parsing.  Just an artefact of having a trailing \n at the end of
the dataset I'm multiplying to get the larger dataset (without it the
original dataset created for x*n, n+1, instead of 2n+1 records, as it would
concatenate the last and first records of each pair).  Haven't noticed the
_leading_ empty record in my tests, so no idea what to do with it.

Looks like C was the answer though :) .  Enjoy yourself,
Mike

-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Darrell
Sent: November 11, 1999 6:18 PM
To: Mike Fletcher; Darrell; python-list at python.org
Subject: Re: ASCII delimited files


----- Original Message -----
From: Mike Fletcher <mcfletch at vrtelecom.com>
> Okay, here's an entirely different approach, extremely memory intensive,
> fairly fast.

I'm impressed if you just whipped this up today!
Should it trim the leading empty record ???
I noticed that it returns an additional empty record.

Here's the 'C' version with your tests.

Time for file of 365 bytes
 (5 * testset, 10 records found):
 0.0
Time for file of 3650 bytes
 (50 * testset, 100 records found):
 0.0
Time for file of 36500 bytes
 (500 * testset, 1000 records found):
 0.0160000324249
Time for file of 365000 bytes
 (5000 * testset, 10000 records found):
 0.15600001812
next test may take 15 seconds or so
Time for file of 3650000 bytes
 (50000 * testset, 100000 records found):
 1.89099991322
next test may take 45 seconds or so
Time for file of 10950000 bytes
 (150000 * testset, 300000 records found):
 7.32799994946

--Darrell


--
http://www.python.org/mailman/listinfo/python-list




From tgfuchs at my-deja.com  Wed Nov 10 10:11:40 1999
From: tgfuchs at my-deja.com (Thomas Fuchs)
Date: Wed, 10 Nov 1999 15:11:40 GMT
Subject: Text Manipulation in Python
References: <VA.00000012.055a8702@ops.corpex.com>
Message-ID: <80c1v7$umd$1@nnrp1.deja.com>

In article <VA.00000012.055a8702 at ops.corpex.com>,
  edward at corpex.com wrote:
> I am completely new to Python.
>
> We want to use it to alter specific lines in template files, typically
> something like changing:-
>
> Variable = 1234
>
> to
>
> Variable = Company Name
>
If you can specify the format of the template file, use Python syntax.
For example:

# file: mytemplate.py
t = '''\
-----------------------
Variable = %(company)s
-----------------------
'''

# file: mymain.py
import mytemplate
company = 'CoCo'
f = open('myoutput', 'w')
f.write(mytemplate.t % locals())
f.close()



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



From timmuddletin at news.vex.net  Sun Nov 28 11:50:45 1999
From: timmuddletin at news.vex.net (tim muddletin)
Date: 28 Nov 1999 16:50:45 GMT
Subject: Pythonwin Reset
References: <01bf38df$8127e0c0$3a42c3d1@Cyberus.cyberus.ca> <_X004.7432$VA6.38559@news-server.bigpond.net.au> <01bf39bd$a28831a0$4155c3d1@Cyberus.cyberus.ca>
Message-ID: <81rmh5$9e$1@hub.org>

On Sun, 28 Nov 1999 16:29:35 GMT, Gordon Williams <g_will at cyberus.ca> wrote:
>NameError: __main__
>
>what has gone wrong?

A non-obvious gotcha. You have to "import __main__" first, to get access
to the namespace in that way.





From bernhard at alpha1.csd.uwm.edu  Wed Nov 10 14:44:16 1999
From: bernhard at alpha1.csd.uwm.edu (Bernhard Reiter)
Date: 10 Nov 1999 19:44:16 GMT
Subject: Text Manipulation in Python
References: <VA.00000012.055a8702@ops.corpex.com>
Message-ID: <slrn82jioa.b4q.bernhard@alpha1.csd.uwm.edu>

On Wed, 10 Nov 1999 08:46:42 GMT, Edward Hasted <edward at corpex.com> wrote:
>I am completely new to Python.
Welcome to a land of programming pleasure.

>We want to use it to alter specific lines in template files, typically 
>something like changing:-
>
>Variable = 1234
>
>Variable = Company Name
>
>The text manipulation strings within Python seem to work sequentially 
>rather than on a line basis.
>
>1. Is this correct?
If I understand you correctly, yes.
But (of course a but) you can relatively easy make python
work on a per line basis.

>2. What is the best way to do text searching and manipulation within 
>Python.

Some other posters suggested looking into other tools, but I would refrain
from that. For one reason python is a unified language and a lot easier to 
learn. The syntax of awk and sed can be tricky in times. And you might 
be able to pull it off with bash scripting or vim replacements rules. Don't.

Use the nice fileinput module to get lines:

-----------
#! /bin/env python
"""" Per line processing example. """"

import sys
import string
import fileinput

def process(line):
	""" Process one inputline and spit out, what we want. """ 
	# get rid of the possible newline and the whitespace around chars
	line=string.strip(line)
	sys.stdout.write(line)

for lines in fileinput.input():
	process(line)
-----------

Now you can use the string module functions to deal with one line,
play with pythons "Slicing" capabilties on the strings or even
pull out the big guns and play the regular expression card with the re module.

The fileinput module can even do in place renaming of files for you.
Oh you call that little program just with the filename(s).

With time you will come to the situation, when you want to pull some
more complicated tricks. Here is a rough example in which I tried
to imitate the way awk enfolds power on all sort of textprocessing tasks.
Don't read it right now, but look into it later... if you feel like it.
it contains a bunch of python related convetions and some variables not
used in this example, but ready for other textprocessing tasks.


Input files look like:

### Opening logfile (channel #whatever), [Mon Nov  2 09:22:49 1999]
[Mon Nov  2 09:23:04 1999]<willi7,0;> Gut.
[Mon Nov  2 09:23:11 1999]<willi7,0;> Wer leitet ? 
[Mon Nov  2 09:23:37 1999]<Bob7,0;> immer der Protokolleur
[Mon Nov  2 09:23:37 1999]#intevation> Lasst mich doch leiten, dann kann jemand anders Protokoll schreiben. :)
[Mon Nov  2 09:24:19 1999]<willi7,0;> Scheint ja gekl?rt. 

Outputfiles look like: 

### Opening logfile (channel #whatever), [Mon Nov  2 09:22:49 1999]
[09:23:04] <willi> Gut.
[09:23:11] <willi> Wer leitet ? 
[09:23:37] <Bob> immer der Protokolleur
[09:23:37] <Cliff> Lasst mich doch leiten, dann kann jemand anders Protokoll schreiben. :)
[09:24:19] <willi> Scheint ja gekl?rt. 


-----------
#! /bin/env python
"""Format and clean tirc normal logfile for IRC session logs.  v%(version)s

USAGE:
 %(progname)s [inputfilename(s)] 

Inputfiles will be replaced, if filenames are given.

The default is to use <Cliff> as user.
The time entries on each line will be cutted to include only the time
and not the day or year, which easily can be seen from the beginning
and end of a consecutive session.

Of course this is dependend on the time format tirc uses.
"""
__version__="0.1"
# initial 2.6.1999 Bernhard Reiter


import fileinput
import sys
import string
import re


nick="Cliff"

def process(line,fileinputobject,write):
    """Process one inputline and spit out what is wanted."""

    # share some things between subsequent calls
    global channel
    global nickre

    #f=string.split(line,";")
    #f=map(string.strip,f)

    # imitate gawk a bit   ;-)
    #NF=len(f)
    FILENAME=fileinputobject.filename()
    NR=fileinputobject.lineno()
    FNR=fileinputobject.filelineno()


    if line[0:3]=="###":
	# opening or closing of logfile

	# grab channelname
	matchobj=re.search('\(channel (#[^)]+)\)',line)
	channel=matchobj.group(1)
	sys.stderr.write("Found channel: " + channel + "\n" )
	
	# prepare nick replacement procedure
	nickre=re.compile(re.escape("]" +channel+">"))
	
    if line[0]=="[":
	# normal line

	# replace funny control string
	line=string.replace(line,chr(3)+"7,0;","",1)
	line=nickre.sub("]<"+nick+">",line,1)
	
	# distance between time and normal text
	if line[25:27]=="]<":
		line=line[:26]+" "+line[26:]

	# cut date and year out 
	line="["+line[12:20]+"]"+line[26:]

    write(line)
	


def main():

    if len(sys.argv)==1:
	sys.stderr.write(__doc__ % 
	    {"progname":sys.argv[0], "version":__version__})

	if sys.platform=="win32":
	    import msvcrt
	    msvcrt.getch()
	    sys.stderr.write("\nPress any key to start reading from stdin.\n")
	else:
	    sys.stderr.write("Now reading from stdin.\n")

    fileinputobject=fileinput.input(sys.argv[1:],1,".org")
    outputwritefunction=sys.stdout.write

    try:
	for line in fileinputobject:
	    process(line,fileinputobject,sys.stdout.write)
    finally:
    	fileinputobject.close()
	

if __name__=="__main__":
	main()
-----------

<cranking-out-python-with-hopes-not-to-confuse-you-too-much>ly,
	Bernhard

-- 
Research Assistant, Geog Dept UM-Milwaukee, USA.    (www.uwm.edu/~bernhard)
Free Software Projects and Consulting 			   (intevation.net)  
Association for a Free Informational Infrastructure              (ffii.org)



From gerrit.holl at pobox.com  Fri Nov 12 14:12:59 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Fri, 12 Nov 1999 20:12:59 +0100
Subject: telnetten vanuit een script
In-Reply-To: <Pine.LNX.4.10.9911121516030.28857-100000@semios.c75.net>
References: <Pine.LNX.4.10.9911121516030.28857-100000@semios.c75.net>
Message-ID: <19991112201259.A6212@optiplex.palga.uucp>

Bert Frederiks schreef:
> Ik wil graag vanuit een script telnetten, dan een paar dingen doen en weer
> uitloggen en ondertussen de uitvoer opvangen in een bestandje. Dat werkt
> prima met pop3 op poort 110 maar op poort 23 lukt het me niet om in te
> loggen. Het probleem zit hem aan invoerkant want dit werkt wel:
> 
> 	telnet njoi 23 >>data.out
> 
> en dan inloggen en dingen doen vanaf het toetsenbord, waarbij de uitvoer
> naar data.out gaat.  Maar ik wil zoiets als:
> 
> 	telnet njoi 23 <data.in >>data.out &
> 
> met data.in een named pipe of een bestand, en dat werkt niet. 
> 
> Heeft iemand een beter idee?

Telnet moet van een terminal lezen, dus je kan expect gebruiken. Expect is
een scripting taal voor dit soort dingen; er is echter ook een python module,
een c library en waarschijnlijk ook een perl module. Sla de documentatie er
eens op na, waarschijnlijk heeft "man expect" je een heleboel te vertellen...

groeten,
Gerrit.
-- 
nl-mop at nl.linux.org: de ongemodereerde Nederlandse moppen mailinglist.
Stuur voor meer informatie een email:
To: Majordomo at nl.linux.org
Subject: maakt niet uit

subscribe nl-mop



From parkw at better.net  Tue Nov 30 19:14:14 1999
From: parkw at better.net (William Park)
Date: Tue, 30 Nov 1999 19:14:14 -0500
Subject: What's the canonical vi setting for Python indentation
In-Reply-To: <38445B6A.A1283D2E@cs.mu.oz.au>
References: <38445B6A.A1283D2E@cs.mu.oz.au>
Message-ID: <19991130191414.A2060@better.net>

On Wed, Dec 01, 1999 at 10:19:06AM +1100, Chris WRIGHT wrote:
> I've tried to find out the "OK" indentation settings for python with vi,
> and a troll through deja-news found some hints (like don't do indent =
> 4). Yet I seem to also remember some advice to the contrary?? I want
> indentation to be 4 characters, and readable by/ acceptable to emacs
> python-mode, and to pass tabnanny...

I use 'vim', and my ~/.vimrc contains (among other things)
    set smarttab	" use 'shiftwidth' at beginning of line
    set shiftwidth=4	" use 4 spaces for indenting
    " indentation after Python keywords
    autocmd BufRead *.py set smartindent cinwords=if,elif,else,for,while,try,except,finally,def,class

	Yours truly,
	William Park

> 
> I think that this should go in the faq...(Couldn't find it from
> www.python.org search)
> 
> 
> cheers
> and thanks
> 
> chris wright
> -- 
> Dr. Chris Wright
> Deputy Director, Intensive Care Unit
> Monash Medical Centre



From lemiss at my-deja.com  Mon Nov  1 15:11:32 1999
From: lemiss at my-deja.com (G. Lewis)
Date: Mon, 01 Nov 1999 20:11:32 GMT
Subject: commas vs. spaces. vs. the legions of new lines
References: <7vj35t$8p1$1@nnrp1.deja.com> <oqiu3mi59y.fsf@titan.progiciels-bpi.ca>
Message-ID: <7vks5h$i1k$1@nnrp1.deja.com>

<pinard at iro.umontreal.ca> wrote:
> language after the fact.  My only fear is that Python will grow popular
> enough so the sad history of c.l.py repeats here, so I hope that Python
> will spread slowly enough so the spirit gets preserved! :-)

Hehe! I hear you. Nice perspective. Possibly not what Guido wants to
hear though. It is true to a certain extent, I'd agree, that
(especially) in this digital domain the faster something gets big the
faster it degenerates in many ways (we see this phenomenon in many
areas), including attracting pretentous attitudes. Though some small
things which never get very big suffer this problem as well; and so far
Python certainly hasn't fallen victim to that protectionistic and
defensive tendancy, so we can have some hope that as it continues to
quickly grow it will also not suffer the same 'attitude' problems that
others have suffered for whatever reason.

Or at least we can watch out for it and fight against it!

After all, besides being universally famous for Silly Walks, aren't
Pythons best known for their prodigious and aggressively friendly
'hugging' abilities? Something for us to keep in mind, perhaps.
<many winks of all sorts of varying percentages>

This is a bit of a silly thread, making mountains out of molehills!
Oh well. Hugs to all Pythoneers.

...


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



From tim_one at email.msn.com  Fri Nov 26 04:58:33 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Fri, 26 Nov 1999 04:58:33 -0500
Subject: automatic "lotjes trekken"
In-Reply-To: <81kll5$211$1@vvs.superst.iae.nl>
Message-ID: <000201bf37f4$cce6f820$202d153f@tim>

[Carel Fellinger wades thru a mass of Tim's chessboard prose, and
 cleverly deduces that he still doesn't have a usable answer]

[Tim]
> ...
> The next twist in the tale is too involved to explain here:  it
> turns out that it's possible to deduce the result from the numbers of
> ways to place 0, 1, 2, ... and N non-capturing rooks on the "cheese-
> inverted" chessboard obtained by putting cheese on all the empty squares
> and removing the cheese from the cheesy squares.  In most applications,
> there are many more empty squares than cheesy squares, so inverting
> their roles can slash the amount of computation.

[Carel]
> and how then can this be deduced, what simple function are we to apply?

Who said "simple" <wink>?  If H(i) is the # of ways to place i non-capturing
rooks on the inverted chessboard, the number of ways to place N
non-capturing rooks on the original chessboard is the sum of the N+1 terms
of this form, where i ranges from 0 thru N inclusive:

    (-1)**i * H(i) * (N-i)! =

    H(0)*N! - H(1)*(N-1)! + H(2)*(N-2)! - ... + (-1)**N*H(N)

H(0) is 1 by definition (why?  because no other value works <wink>).

By way of non-trivial example, let's go back to the number of derangements
of N things.  Then the inverted chessboard is a degenerate one consisting of
cheese everywhere *except* on the main diagonal.  Since all the empty
squares are in distinct rows and columns, the number of ways to place i
non-capturing rooks on the N free diagonal squares is simply C(N, i), where
C(a, b) is the number of combinations of a things taken b at a time =
a!/b!/(a-b)!.  Plugging H(i)=C(N,i) into the above gives the sum of terms of
the form

    (-1)**i * C(N, i) * (N-i)! =
    (-1)**i * N! / i!

which is N! times the sum of N+1 terms of the form

    (-1)**i / i!

Note that the latter sum consists of the first N+1 terms of the Taylor
series expansion for exp(-1), so is an excellent approximation to 1/e.  And
that's "why" the number of derangements is approximately N!/e (the exact
number is gotten by cutting off the approximation to 1/e at the N+1'st
term).

I haven't explained why any of that *works*, and that's still too involved
to go into here.  Any good text on combinatorics will explain it in detail,
though!  In brief, it's a straightforward application of the "principle of
inclusion/exclusion" (which may be called that, or hiding in a chapter on
"sieve formulas") to the "rook polynomial" of the cheesy squares (which is a
simple flavor of generating function).  You already know where the "rook"
comes from, so the rest is a small matter of filling in details <wink>.

Anyway, forget all that unless it interests you.  The news->mail gateway has
disgorged other old msgs in this thread, from which I deduce you don't have
an application in mind for more than 10 people anyway.  I'll attach
something unrelated to the theory above that simply works, and indeed works
faster the more restrictions there are.

Here's an example showing why you should *not* take the other approaches
that have been suggested:

>>> b = Board()
>>> for name in "Carel", "Guido", "Martijn", "Tim":
        b.register_name(name)

>>> b.diagonalize()  # force derangement
>>> b.place_cheese("Carel", "Guido")   # Guido has enough wooden shoes
>>> b.place_cheese("Martijn", "Guido") # ditto
>>> for solution in b.solve():
        print solution

[('Carel', 'Martijn'), ('Guido', 'Carel'),
 ('Martijn', 'Tim'),   ('Tim', 'Guido')]
[('Carel', 'Martijn'), ('Guido', 'Tim'),
 ('Martijn', 'Carel'), ('Tim', 'Guido')]
[('Carel', 'Tim'),     ('Guido', 'Martijn'),
 ('Martijn', 'Carel'), ('Tim', 'Guido')]
>>>

That means there are exactly three permutations that avoid the forbidden
positions.  Note that in two of them, you have to tease Martijn, and in only
one do I enjoy the pleasure of your gift-selecting powers.

That leads to a (possibly) surprising observation:  if you draw lots "at
random" one at a time, you do *not* (in general) have an equal chance of
selecting each of the possible solutions!  In the above, if you draw your
victim first, you have an equal chance of picking me or Martijn.  But you
*should* pick Martijn *twice* as often as me if you're to have equal chances
of ending up with each of the 3 solutions!  If you choose Guido's victim
first, there's no such problem, and ditto if you pick my victim first
(albeit for a different reason) -- but you have no way to know that without
analyzing the set of solutions first.  So what's required isn't backtracking
but forwardtracking <wink>.

Here are three approaches that work:

1) A variant of the one you're already using:  generate a *permutation* at
random, in one gulp.  Then check whether it meets the restrictions.  If it
doesn't, try again.  Then each solution clearly has an equal chance of
getting picked.

Deep flaw #1:  As explained here a few months ago, Python's random-number
generator isn't capable of generating more than about 1E12 distinct
permutations, so even for a list of size 20 it's incapable of generating
"almost all" of the possible permutations.

Deep flaw #2:  As with your current approach, it will run "forever" if there
is no solution.  A variant of the birthday paradox kicks in here too:  even
if all permutations could be generated at random, "at random" means you get
duplicates.  The expected number of permutations you have to generate at
random before you see each of the N! possibilites at least once is about N!
* ln(N!) (ln == log to the base e).  Cut off the search earlier than that,
and chances are better than half that you never saw at least one permutation
(which may have been a solution).

2) The approach below:  generate all possible solutions, then simply apply
random.choice to the list of results.  For your size of problem, it will run
quickly, and in any case runs faster the more restrictions there are.

Deep flaw:  For larger problems, it's utterly impractical (too many
solutions).

3) Go back to the theory:  at each step, analyze the board according to the
method sketched yesterday, and put a rook on that step's distinguised square
with probability given by the number of solutions that have a rook on that
square divided by the total number of solutions (i.e., with or without a
rook on that square).  In the example above, this means that if your first
choice is wrt the Carel->Martijn square, you would first compute that there
are 2 solutions with a rook there and 1 without, and so you should put the
rook there with probability 2/3.  That's the way to do this properly without
backtracking -- and if you think it's worth the effort, your winter is far
too long <wink -- but it is a difficult programming problem!>.

van-lotje-getikt-ly y'rs  - tim

class Board:
    def __init__(self):
        # list of names
        self.names = []

        # map name to index in self.names
        self.name2i = {}

        # set of forbidden positions (pairs of ints)
        self.forbidden = {}

    def register_name(self, name):
        if not self.name2i.has_key(name):
            self.name2i[name] = len(self.names)
            self.names.append(name)

    # Remember that name x can't map to name y.
    def place_cheese(self, x, y):
        for name in x, y:
            if not self.name2i.has_key(name):
                raise ValueError("I don't know " + name + "!")
        self.forbidden[self.name2i[x], self.name2i[y]] = 1

    # Put cheese along the main diagonal (that is, don't let
    # anyone map to themself).
    def diagonalize(self):
        for i in range(len(self.names)):
            self.forbidden[i, i] = 1

    # Return list of all solutions.  Each solution is a list
    # of name pairs.
    def solve(self):
        self.board = board = []
        n = len(self.names)
        rangen = range(n)

        # Set board[i] to a list of its non-forbidden columns.
        cheesy = self.forbidden.has_key
        for row in rangen:
            goodcols = []
            for col in rangen:
                if not cheesy((row, col)):
                    goodcols.append(col)
            board.append(goodcols)
        results = []

        # Build set of all column indices
        columns = {}
        for i in rangen:
            columns[i] = 1

        # Find all solutions.
        _solve(board, 0, columns, [None]*n, results)

        # Translate results back to name pairs.
        names = self.names
        for i in xrange(len(results)):
            asolution = results[i]
            for row in rangen:
                col = asolution[row]
                asolution[row] = (names[row], names[col])
        return results

def _solve(board, row, cols, current, results):
    # board is the current board.
    # row is the index of the row to fill on this call,
    # and increases by one for each recursive call.
    # cols is the set of column indices still available.
    # current is a list whose i'th entry is the index of the
    # column to which row i maps (current solution so far).
    # results is the list of all complete solutions so far.
    if not cols:
        results.append(current[:])
        return
    # Try all ways of placing in this row (the set of columns not
    # forbidden to this row intersected with the set of columns not
    # already filled by shallower calls).
    still_available = cols.has_key
    for col in board[row]:
        if still_available(col):
            current[row] = col
            del cols[col]
            _solve(board, row+1, cols, current, results)
            cols[col] = 1





From gunesh at hotmail.com  Tue Nov 30 06:10:38 1999
From: gunesh at hotmail.com (ganesh)
Date: Tue, 30 Nov 1999 16:40:38 +0530
Subject: Does Pythonwin support SDI. - A Newcomer to Pythonwin
Message-ID: <821fmv$4b5$1@news.vsnl.net.in>

I understand that pythonwin is an MDI application by itself. But I am trying
to develop a SDI application in Pythonwin and have no idea where to start as
there is no proper documentation or PyCSingleDocTemplate object in
Pythonwin. Can anyone help me out.






From gawron at obop.com.pl  Fri Nov  5 04:52:25 1999
From: gawron at obop.com.pl (=?iso-8859-2?Q?Przemys=B3aw?= G. =?iso-8859-2?Q?Gawro=F1ski?=)
Date: Fri, 05 Nov 1999 10:52:25 +0100
Subject: rs232
Message-ID: <3822A8D8.8093F56D@obop.com.pl>

Hello, I'm very new to Python but I'm having a great time learning it !

What is the simplest way to read from the RS232 port under windows,
maybe there is a module for that ready ?

Thanks

Przemek

--
Przemyslaw G. Gawronski UIN:8358522
mailto:gawronskip at usa.net
mailto:gawron at obop.com.pl





From u970130 at studbo.hit.no  Mon Nov  1 21:06:12 1999
From: u970130 at studbo.hit.no (Thomas Weholt)
Date: Tue, 02 Nov 1999 02:06:12 GMT
Subject: MPEG Handling Classes?
References: <381BB206.147DF61C@Flash.Net>
Message-ID: <381e459f.2480674@news.eunet.no>

On Sun, 31 Oct 1999 03:09:43 GMT, Luis Cortes <LCortes at Flash.Net>
wrote:

>Hello,
>
>    Just wanted to know if there are any MPEG handling classes  and
>where I might find them.
>
>Please reply to sender,
>Thanks,
>Luis.
>

I hassled alot of people about this earlier, so you can scan thru
earlier postings, using stuff like ww.deja.com etc.

What I found out is that Guido himself has a module ready, but cannot
release it for licencing reasons ( :-< ) . There is another module
available, which is referred to in postings mentioned earlier. It
works for the basic stuff, but bitrate is not supported ( PST from
www.pythonware.com might get it though ) and running time, sampling
rate etc. is not supported.

Guido said he hesitated releasing his module due to the number of
people wanting it ( he said only me and another guy asked for it . ) 
Still, hope he gets it out somehow. I really need all that other
stuff, not just artist, title, genre etc. that the module available
now support.

--
Thomas Weholt



From wuhy at tcts.seed.net.tw  Sun Nov 14 09:03:01 1999
From: wuhy at tcts.seed.net.tw (Mark Wu)
Date: 14 Nov 1999 14:03:01 GMT
Subject: [HELP] TKinter could not run in Python!
References: <80i63e$7t4$1@news.seed.net.tw> <memo.19991113182523.24605A@forestfield.cix.co.uk>
Message-ID: <80mfel$p3k$1@news.seed.net.tw>

Dear David:
    I think I already solve this problem. Because I have to use TCL/TK to
develop software, therefore I have to keep the TCL/TK in the update to date
version.
    I trying to uninstall the TCL/TK v8.1 , then reinstall the TCL/TK v8.05,
then I make some modification of my env. and TCL source

1) I set TCL_LIBRARY="the path where you install the TCL lib"
2) then, set TK_LIBRARY="the path where you install the TK lib"
3) Finally, I make some modification of some scripts in the 8.05 version...
4) It works ...

Thanks, I hope could share this experience with you!

Best Regards, Mark


David F Hughes <david at forestfield.co.uk> wrote in message
news:memo.19991113182523.24605A at forestfield.cix.co.uk...
> In article <80i63e$7t4$1 at news.seed.net.tw>, wuhy at tcts.seed.net.tw (Mark
> Wu) wrote:
>
> > TclError: Can't find a usable init.tcl in the following directories:
> >     {d:\progra~1\tcl\lib\tcl8.1} {} ./lib/tcl8.0
> > D:/Python/tcl8.0/library
> > D:/Pyt
> > hon/Tools/library
> >
>   <Snip>
>
> >
> > This probably means that Tcl wasn't installed properly.
>
>
> As another Python newbie, I have just spent a couple of hours grappling
> with the same problem. I tracked it down to copies of Tcl80.dll and
> Tk80.dll that had gotten installed in [Windir]\system. Removing them
> solved the problem.
>
> It was also necessary to tell Python where to find the real Tcl
> installation by adding C:\Progra~1\tcl\bin to the PATH command in
> autoexec.bat - I think this would have been done automatically during the
> Python install, but it had been confused by the above.
>
> I believe that the dll's had been put into the \system directory by a
> (rather garish)Tcl GUI builder called SpecTcl (from SUN microsystems dated
> September 1997) which I had installed to evaluate but decided not to use -
> well anyway, SpecTcl stopped working after removing them!
>
> Regards,
> David Hughes
> Forestfield Software Ltd.
> Horsham, UK.
> [david at forestfield.co.uk]





From jlj at cfdrc.com  Fri Nov 19 11:06:08 1999
From: jlj at cfdrc.com (Lyle Johnson)
Date: Fri, 19 Nov 1999 16:06:08 GMT
Subject: FXPy-0.99.87 - a Python interface to the FOX GUI library
Message-ID: <38357510.331300274@192.77.87.74>

This is an update of FXPy, a Python extension module which provides
an interface to the FOX cross-platform GUI library:

  Home Page:        http://home.hiwaay.net/~johnson2/FXPy
  Download Source:  ftp://ftp.cfdrc.com/pub/FOX/FXPy-0.99.87.tar.gz
  Windows Binaries: ftp://ftp.cfdrc.com/pub/FOX/FXPy-0.99.87-win32.exe

To build FXPy from the sources, you will also need to download version
0.99.87 of the FOX library:

  FOX Home Page:    http://www.cfdrc.com/FOX/fox.html
  Download Source:  ftp://ftp.cfdrc.com/pub/FOX/fox-0.99.87.tar.gz
                    ftp://ftp.cfdrc.com/pub/FOX/fox-0.99.87.zip

As you may have guessed by now, the latest release of FXPy is
version 0.99.87.

What is it?
-----------

FXPy is a Python extension module which provides an interface to the
FOX
cross-platform GUI library. With a few minor exceptions, FXPy provides
a
complete interface to FOX. FOX is a C++-based toolkit for developing
graphical user interfaces easily and effectively. Some of the
significant features of FOX include:

 *  A rich set of widgets (including dials, shutters, tree lists, and
    many other "modern" widgets).
 *  Powerful but easy-to-use layout managers.
 *  Extensive support for 3-D modeling using OpenGL or Mesa.
 *  Supports the XDND protocol for drag-and-drop.
 *  Registry for persistent application settings.
 *  Runs natively under Unix/X and Microsoft Windows.
 *  And much, much more!

Please see the FOX home page (URL listed above) for more details about
the FOX library's features. Also consider subscribing to the fox-users
mailing list for the latest news about FOX development; again, see the
FOX home page for information on this.

FXPy is (c) 1999 Lyle Johnson (jlj at cfdrc.com) and is released under
the
GNU Lesser General Public License.

<P><A HREF="http://home.hiwaay.net/~johnson2/FXPy">FXPy-0.99.87</A> -
interface to the FOX cross-platform GUI library.  (19-Nov-99)




From malcolmt at smart.net.au  Fri Nov  5 05:55:04 1999
From: malcolmt at smart.net.au (Malcolm Tredinnick)
Date: Fri, 5 Nov 1999 21:55:04 +1100
Subject: Apply a function to each list member?
In-Reply-To: <XFMail.991105110617.mikael@isy.liu.se>; from Mikael Olofsson on Fri, Nov 05, 1999 at 11:06:17AM +0200
References: <slrn824gee.1f7.bwinton@tor.dhs.org> <XFMail.991105110617.mikael@isy.liu.se>
Message-ID: <19991105215504.A497@Ridcully.home>

On Fri, Nov 05, 1999 at 11:06:17AM +0200, Mikael Olofsson wrote:
> I guess the terminology originally is taken from mathematics. A map is 
> a mathematical concept, similar to a function. Perhaps you should say
> that a function is a map, rather than the other way around.

Correct. In fact, since we're being precise, let me point out that
(mathematically speaking), all functions are maps (or mappings), but not
all mappings are functions. The point is that a function should have only a
single "output" for any given input (i.e. it should be many-to-one or
one-to-one, not one-to-many). So some mapping that sends, say, the number 2
to the numbers 3 and 4 is not a function (but a mapping that sends both the
numbers 3 and 4 to the number 2 is a function, since for each input there
is only a single output).

(Hmm .. I've just realised that a few of my recent postings have had
amazingly little to do with Python. I promise to stick to the topic from
now one. Swear! :-) )

Cheers,
Malcolm Tredinnick




From ionel at psy.uva.nl  Wed Nov 24 13:47:20 1999
From: ionel at psy.uva.nl (Ionel Simionescu)
Date: Wed, 24 Nov 1999 19:47:20 +0100
Subject: Absolute newbie question
References: <Pine.SOL.3.96.991124165729.11616B-100000@giasbma>
Message-ID: <81hbol$5ng@mail.psy.uva.nl>



Check:

http://www.python.org/doc/



Philip Tellis <tellis at giasbma.vsnl.net.in> wrote in message
news:Pine.SOL.3.96.991124165729.11616B-100000 at giasbma...
| Where are the man pages for python?  I am new to python, and would like to
| look through the man pages before I start.  Someone please help.
|
| Philip
|
|





From boud at rempt.xs4all.nl  Tue Nov 30 01:57:06 1999
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 30 Nov 1999 06:57:06 GMT
Subject: XPython (X11)
References: <383E5350.1E07536A@xrce.xerox.com> <slrn83tj53.k7h.bernhard@alpha1.csd.uwm.edu> <81ms2n$126$1@essle.valdyas.org> <slrn840dt7.mov.bernhard@alpha1.csd.uwm.edu> <m3hfi64oyh.fsf@phaedrus.zen.xom> <slrn845im9.178.bernhard@alpha1.csd.uwm.edu>
Message-ID: <81vsg2$g8$2@essle.valdyas.org>

Bernhard Reiter <bernhard at alpha1.csd.uwm.edu> wrote:

> True. But I want that over an integrated API and GNOME and KDE
> don't offer that yet. And they also don't offer it cross platform.
> wxWindows at least tries to get there.

I'm curious as to what you mean by an 'integrated API' - just
curiosity.

>>And yeah, last I heard you had to buy the Troll Kit for PyQt under
>>Win32.  
> Dropping platform compatibility, that a huge point.

Oh well, platform compatibility that costs money is still
platform compatibility.

-- 

Boudewijn Rempt  | http://denden.conlang.org



From mwh21 at cam.ac.uk  Wed Nov 17 13:14:10 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 17 Nov 1999 18:14:10 +0000
Subject: Python equivalent to the perl `backtick` operator?
References: <d0BY3.191219$5r2.437102@tor-nn1.netcom.ca>
Message-ID: <m3d7t90yvh.fsf@atrus.jesus.cam.ac.uk>

"Kevin Howe" <khowe at performance-net.com> writes:

> How do you execute a program from python?
> 
> In Perl you would do it like this:
> 
>     $varName = `programName`;

import commands
varName = commands.getoutput("programName")

- or probably better -
 
result, varName  = commands.getstatusoutput("programName")

HTH,
Michael




From tismer at appliedbiometrics.com  Tue Nov 16 15:28:07 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Tue, 16 Nov 1999 21:28:07 +0100
Subject: list.without()?
References: <NDBBIAPHEKGOBIHCBLIIEEOOCAAA.mcfletch@vrtelecom.com>
Message-ID: <3831BE57.366ED454@appliedbiometrics.com>


Mike Fletcher wrote:
> 
> Hmm, actually, it's only cheap for certain subsets of the problem (which I
> took to be the common case).  In particular, it shines for small numbers of
> instances in lists of fairly arbitrary size.  It falls down completely when
> you are removing large numbers of instances (such as all of them).  I just
> tested five algos, testing 1 instance in middle of list, and a list composed
> of only elements to be removed, findings, module, and raw test results
> below...

Oh well, I overlooked that you have the try inside the while.
That's bad. I thought I had read this:

def without( source, element):
        temp = source[:]
        try:
                while temp:
                        temp.remove( element )
        except:
            return temp
        return temp

Did you try this (I didn't, have no time).

cheers - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From gmcm at hypernet.com  Mon Nov 29 08:05:01 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Mon, 29 Nov 1999 08:05:01 -0500
Subject: Python complaints
In-Reply-To: <FLxz96.IEI@world.std.com>
Message-ID: <1268241794-29392927@hypernet.com>

Will Ware wrote:

> I find myself grumbling about having to type "x = x + 1". The
> really clean thing to do, given that integers are objects, would
> be to define increment and decrement methods, so you'd type
> something like "i.incr()". Then you'd want to go around looking
> for other objects for which it made sense to increment and
> decrement. Long ints and floats can do what integers do.
> Sequences (lists, tuples, strings, xranges) might grow and shrink
> in length. No good idea what to do for dictionaries. -- 

Now that you've had some coffee this morning, do you see 
any problems with this, Will? I mean, assuming you don't 
really want 0 == 1 to be sometimes true?

Seems to me that "++" would have to be an operator that, for 
immutable operands, magically rebound the operand; while 
calling an __incr__ method on mutable objects.

- Gordon



From parkw at better.net  Sun Nov  7 11:46:36 1999
From: parkw at better.net (William Park)
Date: Sun, 7 Nov 1999 11:46:36 -0500
Subject: Python and generating images for web
In-Reply-To: <3824F7AC.4B1BC7F6@yahoo.com>
References: <3824F7AC.4B1BC7F6@yahoo.com>
Message-ID: <19991107114636.A4281@better.net>

On Sat, Nov 06, 1999 at 10:53:16PM -0500, Alexander Stepanov wrote:
> Hello,
> 
> 
> I am a newcomer to web programming and python.  I need to create a
> small web application to make some charts based on user's data input.
> 
> Is it possible to generate gif or png images on a fly using Python?
> 
> I will appreciate if you point me out accessible chart/plot  modules
> and libraries as well as documentation and working sites.    Any
> additional information is strongly welcome.
> 
> (I am in a process of comparison Perl and Python to solve this
> problem.  I plan to use the application under WindowsNT4.0 environment
> + Perl? or Python interp-r).
> 
> Sincerely,  Alexander email: a_stepanov at yahoo.com

I don't know about others, but I use 
    1. Python script to generate data plot in PiCTeX format
    2. TeX and 'dvips -E' to produce EPS file
    3. ImageMagick 'convert' to convert to GIF file
All the steps can be called from Python script, although I usually do
them manually.

	Yours truly,
	William Park



From mikael.johansson at helsinki.xx  Fri Nov 12 10:21:40 1999
From: mikael.johansson at helsinki.xx (Mikael Johansson)
Date: Fri, 12 Nov 1999 17:21:40 +0200
Subject: print "foo", without a space
References: <000b01bf2ccb$70ba3d20$fd2d153f@tim>
Message-ID: <382C3084.C7B01C10@helsinki.xx>

Hello All!

> > Is there another way of getting print to not make a newline than
> > using:
> >     print "foo",  ?
> 
> No.

OK, thanks to all who replied, this was what I really wanted to know.
The other tricks where known to me, but the .softspace thingie seemed
interesting (even if not that usable :-)

The reason for me wanting the print-statement to not make an auto-
space was just to have one simple output-command.

The reason for this again is that I've been assisting in a
thermodynamics course here at the university. I've been responsible
for trying to get the students to do some small proggies (calc.
Gibbs energies etc.), and none of them had any programming experience
to start with, so I chose Python as it seemed (and still seems) a
nice and quite easily adoptable language. I hadn't programmed Python
before either, but now I love it :-)  But anyway, the concept of
stdout and so on are a bit too complicated to get through to the
students in the very limited time I have (it's not supposed to be a
programming course after all). 

Have a nice day,
    Mikael J.



From garry at sage.att.com  Mon Nov 29 10:28:48 1999
From: garry at sage.att.com (Garrett G. Hodgson)
Date: Mon, 29 Nov 1999 15:28:48 GMT
Subject: bug in tkinter canvas event handling?
Message-ID: <38429BB0.1BFFB54B@sage.att.com>

ok, i'm back with another tk problem.
what i want to do is attach a callback to a shape in a canvas
that will tell me which domain level object was picked.

seems simple, but i don't get what i expect.  in particular,
if i press a mouse button while pointing to one object, then drag
to another and release, the release event callback tells me i
was pointing to the first object.  i've distilled it down to a
small test case, attached below.  

this seems like a bug to me, but experience has shown that it's
probably a misconception on my part about how the event handling
is supposed to work.


anyone have any ideas?


------------------------- code follows -----------------------

from Tkinter import *

def Up( id ):
        print 'Up:', id

def Down( id ):
        print 'Down:', id

root = Tk()
canvas = Canvas( root )
canvas.pack()

Wid = 40
Hgt = 25
x, y = 100, 64
shape1 = canvas.create_oval( x-Wid, y-Hgt, x+Wid, y+Hgt, fill='red' )

x, y = 200, 64
shape2 = canvas.create_oval( x-Wid, y-Hgt, x+Wid, y+Hgt, fill='blue' )

canvas.tag_bind( shape1, "<1>",                 lambda e: Down( 'shape1'
) )
canvas.tag_bind( shape1, "<ButtonRelease-1>",   lambda e: Up( 'shape1' )
)

canvas.tag_bind( shape2, "<1>",                 lambda e: Down( 'shape2'
) )
canvas.tag_bind( shape2, "<ButtonRelease-1>",   lambda e: Up( 'shape2' )
)

root.mainloop()


-- 
Garry Hodgson			"Hey, mister, can ya tell me,
garry at sage.att.com		where a man might find a bed?"
Software Innovation Services	He just grinned and shook my hand,
AT&T Labs			"No", was all he said.



From flognat at hotmail.com  Wed Nov  3 16:35:54 1999
From: flognat at hotmail.com (Flogg)
Date: Wed, 03 Nov 1999 21:35:54 GMT
Subject: dejasearch.py - an www.deja.com client, Alpha version comments wanted
Message-ID: <7vq9rq$fv9$1@nnrp1.deja.com>

Did anyone try it out, and got it to work??

I tried it outside our firewall/proxy and Deja seemed to behave a
little bit different, so I have updated the archive.

p.s. lets hope I don't mess this up so you see this message 3 times..

        /Andy
--
   The eye of the beholder lies on the beauty.


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



From michel at digicool.com  Tue Nov 30 17:15:30 1999
From: michel at digicool.com (Michel Pelletier)
Date: Tue, 30 Nov 1999 17:15:30 -0500
Subject: [Zope-dev] Re: Exposing COM via XML-RPC or Something Else
Message-ID: <613145F79272D211914B0020AFF64019276318@gandalf.digicool.com>


> -----Original Message-----
> From: Thomas Heller [mailto:thomas.heller at ion-tof.com]
> Sent: Tuesday, November 30, 1999 2:52 PM
> To: Brian Lloyd; 'Edward Muller'; python-list at python.org
> Cc: zope-dev at zope.org
> Subject: [Zope-dev] Re: Exposing COM via XML-RPC or Something Else
> 
> 
> > You might want to take a look at MS' SOAP (Simple Object Access 
> > Protocol) spec - it's not fully baked yet, but it basically takes
> > xml-rpc several steps further, allowing you to communicate with 
> > SOAP-enabled objects (COM or otherwise) using XML over HTTP. We in 
> > Zope-land are keeping a close eye on this, as publishing objects 
> > via HTTP is the core idea of Zope already.
> 
> Zope and SOAP sounds very similar.
> Is this by accident?

Yup

-Michel



From mcfletch at vrtelecom.com  Wed Nov 10 12:49:30 1999
From: mcfletch at vrtelecom.com (Mike Fletcher)
Date: Wed, 10 Nov 1999 12:49:30 -0500
Subject: Python company worth $100M
In-Reply-To: <80c97t$r6$1@nntp9.atl.mindspring.net>
Message-ID: <NDBBIAPHEKGOBIHCBLIIMEOECAAA.mcfletch@vrtelecom.com>

Nope, don't think you're missing anything. If I recall correctly, the
asyncore docs even mention their assumption that the processing to be done
per call is trivial.  Since all other channels are "blocked" while the
compute-heavy call is processing, the system will suffer performance
degradation for all other clients.  Thread-pools seem to be a decent
approach for longer-running calls.

Cheers,
Mike

-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Aahz Maruch
Sent: November 10, 1999 12:16 PM
To: python-list at python.org
Subject: Re: Python company worth $100M


In article <033201bf2b73$ad1e49e0$f29b12c2 at secret.pythonware.com>,
Fredrik Lundh <fredrik at pythonware.com> wrote:
>
>  [ Searchbutton.com ]
>and if you look at the technical credits page, what do you
>find?
>
>    http://www.searchbutton.com/misc/technicalCredits
>
>asynclib again, of course ;-)

Actually, we're moving away from asyncore in favor of threads.  Unless
we're missing something obvious, asyncore just doesn't do well with
compute-heavy processes.
--
> I wrote:
> > a slightly hairer design issue is what combinations
> > of pattern and string the new 're' will handle.
>

--
http://www.python.org/mailman/listinfo/python-list




From cramu at my-deja.com  Thu Nov  4 06:13:14 1999
From: cramu at my-deja.com (cramu at my-deja.com)
Date: Thu, 04 Nov 1999 11:13:14 GMT
Subject: Template for CGI
References: <7vn0sk$7cp$1@news.fsu.edu>
Message-ID: <7vrpo8$hbu$1@nnrp1.deja.com>

Hi

There is a sophisticated way of doing it. Take a look at
"Generating Dynamic Documents From Templates" from

http://www.python.org/doc/Publications.html

Ramu


In article <7vn0sk$7cp$1 at news.fsu.edu>,
  "Glenn Kidd" <gek8365 at garnet.acns.fsu.edu> wrote:
> I am a newbie to Python and I have a question.  I want a CGI script
that
> will read a file template and replace certain words with variables.
Ii
> tried putting the variable names in the file but it is read as palin
text.
> Is there a way to embed variables into a teplate or a replace function
in
> Python?  Thank you for any help you give.
>
> Glenn Kidd
>
>


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



From news at dorb.com  Thu Nov  4 20:23:56 1999
From: news at dorb.com (Darrell)
Date: Thu, 4 Nov 1999 20:23:56 -0500
Subject: Namespace question...
References: <7vqgib$l8v$1@nnrp1.deja.com> <3820D46C.517EA278@x.x>
Message-ID: <fjqU3.4117$JD1.271306@typhoon.nyroc.rr.com>

Untested code:

sys.modules['__main__'].Version

I'm not sure this is a good idea. If your code becomes part of some greater
application it won't work. A slight improvement might be something like
this.

sys.modules['__keeperOfTheVersion']=sys.modules[__name__]

--
--Darrell
Eric Jacobs <x at x.x> wrote in message news:3820D46C.517EA278 at x.x...
> oofoe at my-deja.com wrote:
> >
> > Hi!
> >
> > How can you access the global namespace in python 1.5.1? I'd like to get
> > a version string which is declared in the main script file. However,
> > each module also has a version string which overrides the original one.
> > I tried to access it as __main__.Version, similarly to $main::Version in
> > Perl, which is what I'm more familiar with. Of course, it didn't work.
> >
> > I have something like this:
> >
> > module file foo.py:
> > ------------------------------
> > Version = '$Id$'
> >
> > def bar():
> >    print "Version is: ", __main__.Version
> >
> > main program:
> > ------------------------------
> > Version = '$Id$'
> >
> > import foo
> >
> > foo.bar()
> >
> > -------------------------------
> >
> > I'd certainly appreciate any pointers. The namespace stuff seems to get
> > a little hazy around issues like this.
>
> Import it first! Use import __main__.
>
> I think some of this confusion arises because there is no global
> namespace in the sense of a area that exists above all modules
> (a good design design, IMO). The word global simply refers to
> variables at the level of the module, as opposed to within a
> function.
>
> --
> http://www.python.org/mailman/listinfo/python-list





From jblazi at netsurf.de  Thu Nov 11 07:15:37 1999
From: jblazi at netsurf.de (Janos Blazi)
Date: Thu, 11 Nov 1999 13:15:37 +0100
Subject: TKinter question
Message-ID: <burn-80ec1e/INN-2.2.1/catlike@broadway.news.is-europe.net>

I was given the advice to look at the Tk Table and tixGrid features
(widgets?) in Tk. Can I use these features in Tkinter and is their use
documented somewhere?

Janos Blazi






From skaller at maxtal.com.au  Thu Nov 25 18:10:03 1999
From: skaller at maxtal.com.au (skaller)
Date: Fri, 26 Nov 1999 10:10:03 +1100
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <38397696@news.xtml.co.uk>
Message-ID: <383DC1CB.371505FF@maxtal.com.au>

Nick Trout wrote:
> 
> | Basically, I'd be really interested to learn: what _don't_ people like
> | about Python

> It would be *really* nice if someone sped up Tkinter (ie. direct tk calls
> rather than through Tcl which is the way I think it works at the moment).
> Its quite a pain if you have a large GUI or you're trying to work with GL.
> Its a shame with this being the defacto GUI and bundled with Python that it
> is so sluggish. It is very usable though.

Viper will support a builtin, high speed GUI interface. 
The current Linux implementation uses GTK.  On my 166MHz Intel
box, it flies. On a real computer, well, who knows :-)
Note that this is so, even though there are FOUR levels of wrapping
around the GTK calls, a C wrapper, then and ocaml wrapper,
then a Viper wrapper, and finally a Python wrapper!

[A windows port would probably use native MS-Windows]

Seriously, Tk is archaic. Forget it. For CPython, wxWindows
or pyGtk would be better choices. For JPython, obviously
Swing is the GUI of choice. 

-- 
John Skaller, mailto:skaller at maxtal.com.au
10/1 Toxteth Rd Glebe NSW 2037 Australia
homepage: http://www.maxtal.com.au/~skaller
voice: 61-2-9660-0850



From Gaetan_Corneau at baan.com  Wed Nov 24 11:37:03 1999
From: Gaetan_Corneau at baan.com (Gaetan Corneau)
Date: Wed, 24 Nov 1999 11:37:03 -0500
Subject: Python and COM: a bug?
Message-ID: <816010E2456BD111A48700805FBBE2EEFDF822@ex-quebec-u1.baan.com>

Bill,

> How is this "O" written?
> Is it an ATL C++ COM object that uses IDispatchImpl? (i.e. a 
> dual interface) Is it a C++ COM object that implements IDispatch on its
own? 
> (not a dual interface)
> Is it a C++ COM object that uses OLEAUT's builtin IDispatch, but just
> doesn't use ATL? (i.e. a dual interface)
> 
> My hunch would either be a bug in Python dispid code, or your 
> COM object has a bug in some custom IDispatch code.

I'm not sure. I know it is implemented using ATL 3.0, but I don't know the
details (and since I'm a newbie in COM/C++/ATL, I can't tell the
difference). I'll ask.

[Half an hour later]  :)

OK. You may well be right about a bug in our code. I was told that the
interface is not dual and that the dispids are created dynamically the first
time a property is used. So maybe there is a problem in this code and that
the fact that I have never encountered the problem in VBScript is pure luck.
I'm still not sure, because a lot of people are using those COM objects in
dozens of VB Scripts and none have reported this kind of problem. But, hey,
I have seen much stranger occurences than that already :)

What I'll do is write a test script in Python that just tries to reproduce
the problem. If I can finally find a way to reproduce it on demand, I'll ask
our resident COM expert to trace the code to see if the problem is on our
side. If its not, maybe I'll have some more clues. One way or another, this
is one potentially very dangerous bug, for us at least, so I'll invest more
time and effort to find what the problem is.



> I also don't think we try to optimize any calls. COM doesn't 
> give one enough
> information to even bother optimizing anything other than 
> caching Name -> DispID mappings.

That's what I meant by "optimization".


> > And sometimes, I get something like that:
> > 
> > >>> Object
> > <COMObject <unknown>>
> > >>> P = Object.Parent # Parent is an attribute that is the 
> object that
> > created the current object
> > >>> P
> > <method CDispatch.Parent of CDispatch instance at c5e780>
> > >>> P = Object.get_Parent()
> > >>> P
> > <COMObject <unknown>>
> > >>>
> > 
> 
> What really helps in cases like this is to have the IDL for your
> dispatch/dual interface so that we can examine whats 
> happening much easier.
> :)

I put it in attachment.


> 
> The otherthing to note is that the <COMObject <unknown>> 
> occurs because you
> aren't wrapping the typelibrary with Python helper classes. Using
> win32com.client.EnsureModule("<typelib guid>") allows you to support a
> couple of features that are impossible to deterimine without 
> the typelib. I forget exactly what those are off the top of my head.

That's something I don't know about. My (naive) question is: why doesn't
VBScript require such a step?

 
> The answer believe it or not is, "Python is the most powerful 
> scripting > language on the planet for Win32 programming." its multi
threaded, COM
> happy, ActiveX Scripting Engine happy, and really easy to 
> talk to. VBScript
> on the other hand... Don't get me started... :) 

You certainly don't have to convince *me* of that :)
Just the fact that I can develop my Python scripts to test those COM objects
from an INTERACTIVE COMMAND LINE is a *major* feature. I can play with the
objects in a Korn Shell interactively to find the recipe, instead of the
usual edit/save/run/debug cycle.

 
Thanks a lot,
______________________________________________________
   Gaetan Corneau
   Software Developer (Quality Assurance Team)
   BaaN  Supply Chain Solutions  
   http://www.baan.com
   E-mail: Gaetan_Corneau at baan.com                  
   Tel: (418) 266-8252          
______________________________________________________
"Profanity is the one language all programmers know best"





-------------- next part --------------
A non-text attachment was scrubbed...
Name: ITKCOMAPI.idl
Type: application/octet-stream
Size: 26334 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/19991124/2e5a20eb/attachment.obj>

From mlh at vier.idi.ntnu.no  Tue Nov  2 10:27:49 1999
From: mlh at vier.idi.ntnu.no (Magnus L. Hetland)
Date: 02 Nov 1999 16:27:49 +0100
Subject: Drawing (directed) graphs
References: <ye1bt9das2p.fsf@ex.ac.uk>
Message-ID: <y0jzowwsykq.fsf@vier.idi.ntnu.no>

R.M.Everson at ex.ac.uk (R.M.Everson) writes:

> Hello,
> 
> Does anyone know of Python code for drawing graphs of the
> graph-theoretic type? Ideally, I'd like to be able to interactively
> add nodes and edges etc and add data to structures associated with the
> nodes and edges.

OK - so this isn't really an answer to your question... (I.e. it is
not a Python-implementation)... But Graphviz is really good IMO...

  http://www.research.att.com/sw/tools/graphviz/

I'm sure you could SWIG it and script it from Python...
And - maybe we should have something like this for PIDDLE too?

> 
> Many thanks
> Richard Everson.
--

  Magnus          Echelon jamming noise:
  Lie             FBI CIA NSA Handgun Assault Bomb Drug Terrorism
  Hetland         Special Forces Delta Force AK47 Hillary Clinton 



From cgw at fnal.gov  Wed Nov 10 11:54:10 1999
From: cgw at fnal.gov (Charles G Waldman)
Date: Wed, 10 Nov 1999 17:54:10 +0100 (MET)
Subject: Intelligent strptime?
In-Reply-To: <19991110171256.A5196@optiplex.palga.uucp>
References: <19991110171256.A5196@optiplex.palga.uucp>
Message-ID: <14377.41778.712189.7874@gargle.gargle.HOWL>

Gerrit Holl writes:
 > 
 > Does there exist an intelligent strptime? "26/11/99", for example, can _only_
 > mean the 26th November of this year, and the chance that "14:11:12" means
 > 14 november 1912 is very small. Is there a module/function that guesses the
 > format? There isn't doubt very often, only for 12/11/99, US is 11 december,
 > Europe is 12 November (isn't it?).

I think that there's an ambuguity between US-style MM/DD and
European-style DD/MM whenever the day of the month is 12 or less.   Of 
course when DD==MM then the ambuguity doesn't matter (e.g. 3/3/99).
But that still leaves 11 days in each month which have this ambiguity, 
which comes to 132 days per year.  So, given a 365-day year, 36% of
the dates are ambiguous, which doesn't make "guessing" look like a
very attractive option to me.




From trekkan at yahoo.com  Sun Nov 28 14:01:30 1999
From: trekkan at yahoo.com (Troy)
Date: Sun, 28 Nov 1999 11:01:30 -0800
Subject: Parsing text
Message-ID: <g1e04.1$O8.108@news.inreach.com>

As you will be able to tell, I am new to Python so please forgive any newbie
things I might do. =)

What I would like to do is take a variable and parse out each of the pieces.
Most commonly delimited by spaces.  Example text:

This is a test

I would like to be able to pull out any of "This", "is", "a", "test" and be
able to drop any element into a variable that I wanted.

I hope this explains it enough.  Thanks for any help you may be able to
provide.

One other thing, is there a way to tell if a substring exists in a string?
in a basic sort of way

if 'abc' isin '123abc456'

Something to that effect anyway.

Thanks for your time!

Troy





From mwh21 at cam.ac.uk  Fri Nov 19 09:18:58 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 19 Nov 1999 14:18:58 +0000
Subject: Beginning programming with Python
References: <sng0ychgwt.fsf@motorola.com> <382B3A53.5816B2@callware.com> <382B3D79.74CD3E0F@email.sps.mot.com> <80p9ds$ppo$1@newshost.accu.uu.nl> <asteroidal-80psfo/INN-2.2.1/bat@broadway.news.is-europe.net> <38317787.19AFDAF5@compaq.com> <812rg4$b6$1@essle.valdyas.org> <00ef01bf3271$10a8bb90$f29b12c2@secret.pythonware.com> <813a3h$jc0$2@newshost.accu.uu.nl>
Message-ID: <m3n1saef8t.fsf@atrus.jesus.cam.ac.uk>

m.faassen at vet.uu.nl (Martijn Faassen) writes:

> Not to mention you couldn't write:
> 
> if foo: print "Foo!"
> 
> without the colons. Not that I'd mind terribly if that usage went
> away though; it is scattered all across the Zope sources which I was
> browsing through last night, and I don't think that construction
> adds to readability. The ':' in my opinion does, though. Even though
> I forget it myself on occasion.

I don't forget it very often, largely because python-mode goes
obviously screwy if you leave it out...

Cheers,
Michael



From tomr at cnri.reston.va.us  Wed Nov 24 07:29:05 1999
From: tomr at cnri.reston.va.us (Tom Reinhardt)
Date: Wed, 24 Nov 1999 07:29:05 -0500
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <38397696@news.xtml.co.uk> <221119992150405215%zippy@cs.brandeis.edu> <383a8909@news.xtml.co.uk>
Message-ID: <383BDA11.5A479E60@cnri.reston.va.us>

Nick Trout wrote:
> 
> How do you write large/complicated applications in Lisp, the syntax with all
> those brackets drives me mad!
> 
I can't resist this one: From someone who had written a few, much easier
than you might think. Lisp programmers have always enjoyed rich
programming environments, hosting features that are still lacking in
many so-called modern languages. Good type checking,
syntax/language-directed editors, tightly coupled debuggers, etc., were
available before the first CADR (gasp!) machines in the eighties. In
fact, Lisp is syntactically regular making it less taxing than C++ or
even Python (still haven't got the hang of when to use ":"'s and
backspaces) for extended efforts. Finally, most modern Lisp dialects
provide a rich (meta)object substrate as well as GUIs.

> I've never heard of this clever type checking. It sounds like a very good
> idea. It would be nice to be able to choose whether to have typing on or
> off. I know you can run a Lint checker over Python but whats a compiler
> for?!

To keep you honest, and to allow you to design the algorithm independent
of machine or micro(source level) optimization concerns.
> 
> N
> 
> "Patrick Tufts" <zippy at cs.brandeis.edu> wrote in message
> news:221119992150405215%zippy at cs.brandeis.edu...
> | In article <38397696 at news.xtml.co.uk>, Nick Trout
> | <nick at spam_me_notvideosystem.co.uk> wrote:
> |
> | > | Basically, I'd be really interested to learn: what _don't_ people like
> | > | about Python
> | >
> | > The lack of typing is a problem sometimes but a real bonus when writing
> | > stuff quickly. I think speed of execution is my main complaint but I
> believe
> | > that is being addressed in the next version (ie. a jump table for
> bytecode
> | > execution was mentioned etc.).
> |
> | Python could follow the Common Lisp model where variables can be typed
> | or not, and the compiler is welcome to actually use the information if
> | it feels like it.
> |
> | The best current implementation of Common Lisp for type-declarations
> | and doing smart things with them is CMU Common Lisp.  CMU CL has a
> | compiler called ... Python ... which does some amazing things with both
> | explicit and IMPLICIT type declarations.
> |
> | For the latter, the compiler can often prove what the type must be even
> | where there are no declarations, and then use the type-specific numeric
> | operators to improve the efficiency of the code while maintaining
> | complete code safety.
> |
> | Here's an example:
> |
> | (defun foo (x) (mod x 4))    ; return x mod 4, always an integer
> |
> | (defun bar (x) (^ 2 x))      ; return 2^x, always an integer
> |
> | (defun baz (a b) (+ (foo a) (bar b)))  ; foo(a) + bar(b)
> |
> | The compiler determines that an in BAZ, an integer add can be used
> | instead of the general, type-checking add.  The compiler increases the
> | speed of the code by eliminating the type-check, and minimizes the
> | memory used by choosing the integer representation for numbers instead
> | of a general numeric representation.
> |
> | If the Python language is going to be extended to allow for type info,
> | I would love to see as much as possible borrowed from the CMU CL Python
> | compiler, as it's one of the finest intelligent compilers I know of.
> | And it has a cool name.
> |
> | More info on CMU CL: http://www.cons.org/cmucl/
> |
> | And a bit on the CMU Python compiler's type-checking (chap 4.5.2 of the
> | CMU CL User Manual by Robert A. MacLachlan, CMU tech report
> | CMU-CS-92-161):
> |
> | http://www.mindspring.com/~rtoy/software/cmu-user/node98.html#SECTION005
> | 52000000000000000
> |
> | --Pat
> |
> | * foo bar and baz can also return bignums.  I'm ignoring this case to
> | make the example clearer.



From hinsen at cnrs-orleans.fr  Thu Nov  4 05:22:15 1999
From: hinsen at cnrs-orleans.fr (Konrad Hinsen)
Date: 04 Nov 1999 11:22:15 +0100
Subject: eff-bot Python book ... revisited
References: <7va9hf$ngi$1@marcie.fissiontech.com> <00fd01bf23cb$e4b6ed30$f29b12c2@secret.pythonware.com> <381C952F.2E8F30EC@callware.com> <99103116022800.00566@quadra.teleo.net>
Message-ID: <m3hfj27e08.fsf@chinon.cnrs-orleans.fr>

Patrick Phalen <python-list at teleo.net> writes:

> I printed the book both sides, using inexpensive paper, on my LaserJet.
> Whenever I do this, the first pass (odd pages only) subjects one
> surface to intense heat, buckling/warping the stack. This makes it
> nearly impossible to feed the stack back through for the second (odd)
> pass without jamming the printer.

It's a problem with cheap paper. Of course more expensive paper isn't
necessarily better, but the paper that is better is usually more
expensive.

> However, this time, I let the stack sit for several days, to allow the
> moisture in the fibers to reequalize (?) -- and it worked.

You can speed it up by keeping it under pressure (a 16-ton weight
will do nicely ;-)
-- 
-------------------------------------------------------------------------------
Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.55.69
Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais
-------------------------------------------------------------------------------



From none at fake.net  Thu Nov 11 01:53:56 1999
From: none at fake.net (me)
Date: Thu, 11 Nov 1999 06:53:56 GMT
Subject: Win32api.... can it create a Windows shortcut (.lnk) file?
Message-ID: <8KtW3.51128$Rx2.608380@news1.rdc1.bc.home.com>

I'm just trying to find an easy to create shortcuts on Windows platforms
through Python... it doesn't seem to be a simple matter though.  I've gone
through all the Win32 docs, and found nothing, and have started writing
(with limited success) a C function to do it.

Is there an easy way?  q:]

Thanks,
Kevin Cazabon
kcazabon at home.com





From aa8vb at yahoo.com  Mon Nov  1 12:17:09 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Mon, 1 Nov 1999 12:17:09 -0500
Subject: __str__ vs. __repr__
Message-ID: <19991101121709.A134039715@vislab.epa.gov>

1) Is there a convention for what __str__ and __repr__ should return for
   classes?
2) Or, whatever they return, should they return the same value?
3) If so, why have both in the language?

     Searching the archives yields an example with odd behavior that
suggests they should generally return the same value:

   class c:
       def __str__(self):  return 'foo'
       def __repr__(self): return 'bar'
   
   >>> str(c())
   'foo'
   
   >>> str([c(), c(), c()])
   '[bar, bar, bar]'

-- 
Randall Hopper
aa8vb at yahoo.com



From pinard at iro.umontreal.ca  Tue Nov 30 11:32:52 1999
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 30 Nov 1999 11:32:52 -0500
Subject: Python complaints
In-Reply-To: Dan Schmidt's message of "30 Nov 1999 09:26:11 -0500"
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <wkiu2kqcn0.fsf@turangalila.harmonixmusic.com>
Message-ID: <oqk8n09byj.fsf@titan.progiciels-bpi.ca>

Dan Schmidt <dfan at harmonixmusic.com> writes:

> Mark Carroll <markc at chiark.greenend.org.uk> writes:

> | Basically, I'd be really interested to learn: what _don't_ people
> | like about Python?

> The main thing I find frustrating is that lambdas are not very
> powerful.

So do I.  They are a bit awkward and do not fit so well in the language.
Moreover, they are not really needed.  The solution should be to eradicate
lambda's instead of making them more powerful.  This is easily done here:
by choice, I just never use them.  But I would even prefer not having the
choice, for the benefit of Python being a bit simpler.

> Much of the time that I would use a simple 'map {}' in Perl, I find that
> I have to write a separate function [...]

Writing two more lines is not such a big deal, given that usually, people
are using `lambda's rather sparingly, anyway.  Eradicating `lambda's would
probably never yield a source code bloat.

By the way, I find `map' rather useful, even if I read recently that Guido
somewhat regretted having introduced it in Python.  I guess one could write
his/her own `map' function, it is just simpler for the community that the
order and meaning of arguments is decided once and for all in the library.

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





From xtian at regurgitator.zzn.com  Sat Nov  6 00:52:45 1999
From: xtian at regurgitator.zzn.com (Xtian Muirhead)
Date: Sat, 6 Nov 1999 18:52:45 +1300
Subject: ~ 1 = -2 ???
References: <m2vh7hk1ys.fsf@netmeg.net>
Message-ID: <800fpr$ahh4u$1@titan.xtra.co.nz>

Les Schaffer <godzilla at netmeg.net> wrote in message
news:m2vh7hk1ys.fsf at netmeg.net...
> i am doing some bit operations, and dont understand the following:
>
> >>> ~ 1
> -2
>
> huh???? i would have thought that bitwise invert of 1 would be
>
> 0001 --> 1110 = 13 (base 10)
>
> what marble am i missing?
>
> les schaffer

Integers are stored in twos-complement format. (Which means...)
The top bit of the integer is treated as having a negative value.
This means that FFFFFFFF is equal to
1 + 2 + 4 + 8 + ... + 1073741824 - 2147483648 = -1
So, 1 is 00000001, inverted is FFFFFFFE, which is
2 + 4 + 8 + 16 + ... + 1073741824 - 2147483648 = -2

This means that the largest positive integer value is 2147483647, while
the largest negative value is -2147483648. It's better than having the
top bit just be a sign bit 'cause it avoids confusing -0 problems. (And
it gives us one more value. Yay!)

HTH, HAND
Xtian

xtian at regurgitator.zzn.com

"As I bit into the nectarine, it had a crisp juicyness that was very
refreshing, until I realised that it wasn't a nectarine at all, but a
HUMAN HEAD!!!"





From karanvasudevaNOkaSPAM at usa.net.invalid  Sat Nov  6 05:26:17 1999
From: karanvasudevaNOkaSPAM at usa.net.invalid (Karan Vasudeva)
Date: Sat, 06 Nov 1999 02:26:17 -0800
Subject: exec / execfile namespace problem
Message-ID: <004aa0e3.c3e96526@usw-ex0101-002.remarq.com>

Hi,

I'm trying to do this:
execfile('cool.py')

and this:

import fileinput
chunk = ''
for line in fileinput.input('cool.py'):
   chunk = chunk + line
exec(chunk)

where 'cool.py' is this: (in part)
import httplib
import time
import whrandom
h = httplib.HTTP('')
h.putrequest('GET', '/favicon.ico')

This fails with this error message at the command prompt:
Traceback (innermost last):
  File "testSlave.py", line 1, in ?
    execfile('cool.py')
  File "cool.py", line 8, in ?
    h.putrequest('GET', '/favicon.ico')
  File "httplib.py", line xx, in putrequest
    self.send(str)
  File "httplib.py", line xx, in send
    self.sock.send(str)
AttributeError: sock

(I changed two line numbers to xx, they were different but thats
irrelevant)

I've tried to change self.sock to sock, but that didn't help. Guess
I'm stumbling around in the dark being an absolute newbie at this.
Any help would be appreciated.

Regards,
Karan.



* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!




From init__self at my-deja.com  Mon Nov  8 14:49:28 1999
From: init__self at my-deja.com (init__self at my-deja.com)
Date: Mon, 08 Nov 1999 19:49:28 GMT
Subject: IPC8: Call the 800 number!
Message-ID: <8079g6$fd4$1@nnrp1.deja.com>

There seems to be a problem between the chairs and the keyboards at
the 703-524-6400 Key Bridge Marriott hotel -- the reservation agents
don't seem to know about the conference and were unable to speak to
anybody that did. Therefore they wouldn't honor the $129 rate. I even
tried rep fishing but got the same person again (bah!).

However, the 800 number *does* know about the conference and was more
than happy to take my money. You might want to try them first.


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



From duncan at rcp.co.uk  Mon Nov 15 08:08:14 1999
From: duncan at rcp.co.uk (Duncan Booth)
Date: 15 Nov 1999 13:08:14 GMT
Subject: '==' vs. 'is' behavior
References: <382C2C6A.64A8F2C2@theriver.com> <14380.12036.197083.409565@weyr.cnri.reston.va.us>
Message-ID: <8E7F86267duncanrcpcouk@news.rmplc.co.uk>

fdrake at acm.org (Fred L. Drake, Jr.) wrote in
<14380.12036.197083.409565 at weyr.cnri.reston.va.us>: 
>  It's working just fine!  Note that the "small" integers (through 99) 
>are always reused, so the "is" relationship will hold.  It's not very
>useful, but reusing the integer objects saves a lot of allocation of
>small numbers, and makes real-world code a lot faster.
>
It isn't just small integers though that can be reused. Consider:
>>> a = 999
>>> b = 999
>>> print a is b
0

Put the same lines into a file and run it and it will print 1, or in
interactive mode put the lines into a function definition and again it
prints 1. When compiling anything identical constants are folded
together. 





From fredrik at pythonware.com  Tue Nov 30 03:03:09 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 30 Nov 1999 09:03:09 +0100
Subject: '==' vs. 'is' behavior
References: <382C2C6A.64A8F2C2@theriver.com> <14380.12036.197083.409565@weyr.cnri.reston.va.us> <8E7F86267duncanrcpcouk@news.rmplc.co.uk> <oqemd9dk2g.fsf@titan.progiciels-bpi.ca>
Message-ID: <015301bf3b0a$685a3fc0$f29b12c2@secret.pythonware.com>

Fran?ois Pinard <pinard at iro.umontreal.ca> wrote:
> By the way, is there common wisdom (or rather, maybe, usage-standards-to-be)
> about using `is' instead of `=='?

according to the eff-bot:

    foo is None
    type(foo) is type(bar) # same type
    foo is bar # same instance

use '==' for everything else (unless you know what
you're doing, of course).

</F>




From leef_me at my-deja.com  Tue Nov 23 13:11:25 1999
From: leef_me at my-deja.com (Lee Fletcher)
Date: Tue, 23 Nov 1999 18:11:25 GMT
Subject: trouble with python 1.5.2 install on win98
Message-ID: <81elc8$8i2$1@nnrp1.deja.com>

Hoping for some guidance to my problem w/Python,

I have installed py152.exe on (2) Win98 machines:
1 at work, 1 at home

The installation at home runs. I am having trouble running IDLE at
work. It accesses the hard drive but never becomes a running program.

Starting from the beginning, I have d/l and run py152.exe.
I accepted all default suggestions for installation of python, tk &
tcl.
I tried the "verify that the interpreter works " code on the

     http://www.python.org/download/download_windows.html

page, both examples work

I tried accessing IDLE from the "start menu" but no results.
I rebooted (windows restart) to make sure any pending "setups" were
finished.
I tried accessing IDLE from the "start menu" >Again< but >still< no
results.

I have tried deja & faq searches from the ptyhon help page, but did not
see anything that applies to this.

Please, can you make any suggestions on what I have done wrong?

Thanks,
    Lee



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



From claird at starbase.neosoft.com  Thu Nov 11 06:23:02 1999
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 11 Nov 1999 11:23:02 GMT
Subject: FYI: Python company worth $100M
References: <38295BC2.13E2E31F@digicool.com>
Message-ID: <FDBF3FC2E3C6CFD4.BA55F6701D0F5454.EAAC66C1FCC39FBE@lp.airnews.net>

In article <38295BC2.13E2E31F at digicool.com>,
Paul Everitt  <paul at digicool.com> wrote:
>
>Next time someone claims that Python is marginal...next time that
>someone says nobody uses Python...
>
>eGroups is merging with Onelist (a competitor), and each is valued
>around a million bucks:
>
>  http://www.thestandard.com/articles/display/0,1449,7543,00.html
			.
			.
			.
Most readers will be able to supply the missing
word in "... a hundred million bucks" above that
you accidentally lost, Paul.

Incidentally, Python should play a big role in
news that I expect to emerge in the next month
from two of the biggest hardware and software
companies.  Stay tuned to this space for details.
-- 

Cameron Laird           http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX



From tjreedy at udel.edu  Wed Nov 24 00:15:55 1999
From: tjreedy at udel.edu (Terry Reedy)
Date: 24 Nov 1999 05:15:55 GMT
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <383979F7.25D6@cs.bham.ac.uk>
Message-ID: <81fsab$2b$1@news.udel.edu>

In article <383979F7.25D6 at cs.bham.ac.uk>, S.I.Reynolds at cs.bham.ac.uk 
says...

>Oh yeah! I'd I disklike writing 'self.' eveywhere. Some kind of
>shorthand would be nice. e.g. I'd be nice to write:
>  .x.f()
>instead of:
>  self.x.f()

You are currently welcome to save 3 keystrokes (rather than 4)
by naming the method parameter 's' instead of 'self'.  The latter
is just a convention that many people use -- at least for public
consumption.

TJR




From mlh at vier.idi.ntnu.no  Tue Nov  2 10:12:27 1999
From: mlh at vier.idi.ntnu.no (Magnus L. Hetland)
Date: 02 Nov 1999 16:12:27 +0100
Subject: How can I make my assertions smarter?
References: <381D7087.BF8510E5@webone.com.au> <y0jpuxuryuh.fsf@vier.idi.ntnu.no> <381e4350.91314552@news.usyd.edu.au>
Message-ID: <y0j3dupszac.fsf@vier.idi.ntnu.no>

jesse at student.usyd.edu.au (Jesse Sweeney) writes:

> On 01 Nov 1999 16:55:02 +0100, mlh at vier.idi.ntnu.no (Magnus L. Hetland) wrote:
> Simple? Yes. Works? No. <wink>
> 
> The problem is that pre's locals aren't the same as the locals in the function
> from which pre is being called, so you run into problems like this one:

Oops... I guess I was a bit quick there... ;)

--

  Magnus          Echelon jamming noise:
  Lie             FBI CIA NSA Handgun Assault Bomb Drug Terrorism
  Hetland         Special Forces Delta Force AK47 Hillary Clinton 



From da at ski.org  Mon Nov 29 16:11:41 1999
From: da at ski.org (David Ascher)
Date: Mon, 29 Nov 1999 13:11:41 -0800 (Pacific Standard Time)
Subject: Numeric write-in place facility
In-Reply-To: <81tovm$oe0@mail.psy.uva.nl>
Message-ID: <Pine.WNT.4.04.9911291309450.263-100000@rigoletto.ski.org>

On Mon, 29 Nov 1999, Ionel Simionescu wrote:

> Hi,
> 
> In a post on the wxPython list, Markus Gritsch suggested that array ops such
> as
> a = a+b
> can be done faster by using write-in place:
> Numric.add(a,b,a)
> 
> Playing around I see now that other ufuncs allow a write-in target
> specification as their last argument. For example,
> 
> Numeric.sqrt(a,b)
> 
> will put the result in b (if its dimension and type are Ok).
> 
> The docs do not tell much about this.
> Can anyone confirm/clarify/comment ?

1) Confirm: yes
2) Clarify: well, I think you pretty much summed it up
3) Comment: Watch out for 'slices'.  If you do

a = arange(10)
b = a[::-1]
add(a,b,b)
print b

you might be suprised =)

--david ascher




From asc at bsiag.com  Wed Nov  3 07:17:23 1999
From: asc at bsiag.com (Alex Schroeder)
Date: 03 Nov 1999 13:17:23 +0100
Subject: Download Python Imaging Library Handbook
References: <s81eeb86.053@mail.conservation.state.mo.us>
Message-ID: <ueme7bwh8.fsf@bsiag.com>

"Jeffrey Kunce" <kuncej at mail.conservation.state.mo.us> writes:

> I used websucker.py from the python Tools/webchecker directory.
> python websucker.py http://www.pythonware.com/library/pil/handbook/index.htm

I found a printable (PS and PDF) version of the handbook on the SIG
site for an old version (0.2b3 instead of 1.0):

http://www.python.org/sigs/image-sig/Imaging.html

But I guess I will use websucker.py...  :)

Alex.
-- 
http://www.geocities.com/kensanata/
http://www.bsiag.com/



From cfelling at iae.nl  Wed Nov 24 18:45:07 1999
From: cfelling at iae.nl (Carel Fellinger)
Date: 25 Nov 1999 00:45:07 +0100
Subject: automatic "lotjes trekken"
References: <81cqnv$2ip$1@vvs.superst.iae.nl> <bvwu2mddcnd.fsf@room.mathdom.anu.edu.au>
Message-ID: <81hta3$1p7$1@vvs.superst.iae.nl>

Alex <alex at somewhere.around.here> wrote:

> Hi, Carel.

> This is a fun problem.  Basically, you want to find a set of disjoint
> loops in a digraph whose union contains all of the digraph's vertices.
> How big, and how far from complete, is the graph likely to be?  I don't
> know a lot of graph theory, but it seems likely that the problem of

Neither do I, I'm even not sure I understand what you wrote:)

> counting every such set of loops is NP-complete for a general graph.  At

But you're definitely right here. Though n! always looks to me as a harmless
function, something like "really n", it is almost as dreadfull as n^^n.
Fortunately in my family n is rather small, we are five now and even if the
girls get hooked we won't surpase eight for quite some time yet:) Moreover
the problem space is a little below n!, as one shouldn't draw oneself. The
question was: how much lower given some constrains.

The approach I've taken sofar was to simulate real life, let
everybody draw, and as soon as someone draws one he doesn't want,
throw all the lots back in the hat and try again. [[ well Greg, in my
books this wouldn't be classed as a backtracking solution :]] This will
work given that there is a solution, time and the proper functioning of
random.choice. It will work better if the number of possible solutions
comes closer to n! simply because it will get more likely then that
any full drawing will be a solution too. So the more family members
and the less restrictions, the more likely it succeeds in a few tries.
OTOH, the more tries it takes the happier I am that I didn't had to go
through it in real live:)


class Hat:
    ...
    def hussle(self):
	'''hussle the hat until the drawing is valid'''
	if self.possibilities() == 0:
	    raise 'Even eternity won\'t give a valid hussle'
	self.tries = 0
	while 1:
            if self.tries > 999:
                raise 'I am bored, can\'t find a solution quick enough'
	    hat, self._drawn, self.tries = self._lots[:], [], self.tries+1
	    for drawer in self._lots:
		lot = random.choice(hat)
		if lot in self._blanks[drawer]:
		    break
		hat.remove(lot)
		self._drawn.append(lot)
	    else:
		return

    def possibilities(self):
        '''this should return the size of the anwser space,
        but now it simply returns the number of permutations of the lots'''
	U = reduce(lambda x,y: x*y, range(1, len(self._lots)+1), 1) # len!
	return U


Being mad I just wanted to know in how many ways the drawing requirements
could be fullfilled, cause this would allow a more contrieved approach.
Just immagine all solutions being ordered and given a number (e.g. the order
in which they were to be generated by a straight forward backtracking
algoritme), and a way to determine the number of solutions right to any
point in that ordered solution space. So we know the number of solutions,
and can directly 'draw' one with random.choice(range(nr-of-solutions)).
Now we will use a straightforward backtracking algoritme to generate the
solutions in the before mentioned order. *But* instead of generating all
possibilities we skip those parts of the solution tree in which the drawn
solution can't be found. This would lead us straight to the anwser:)


> I thought of another approach that might solve your problem -- to use a
> randomized algorithm which admits a relatively small but still non-zero
> probability to the assignments you'd prefer to proscribe.  Then your
> daughters could comfort themselves with the thought that they're
> unlikely to have to give/receive presents to/from each other.  For

I could live with this, but they would object fearcefully if it happened.

> instance, you could get everyone to assign a positive number to everyone
> else.  They should assign lower numbers to people they'd rather not give
> presents to (perhaps you don't have to worry too much about who they'd
> rather not receive presents from, since they're not supposed to know,
> anyway).  Say we decide to call the people in your family F={P_1, ...,
> P_n}.  Let G_0=R_0=F.  Here, G stands for givers, R for receivers.
> Construct sets G_i, R_i inductively as follows:

> Choose a person P randomly from G_{i-1}.  They've given you a function
> f:F->{positive numbers}.

> Let T=sum {f(P_l) | P_l in R_{i_1}}  Choose a person Q at random from
> R_{i-1} in such a way that the probability of a person Q' being chosen
> is given by f(Q')/T.

> Let G_i=G_{i-1}\P, R_i=R_{i-1}\Q, and let the corresponding pair be
> (P,Q).  Tell P that they're to give something to Q.

As I said, my mind is blurred, but I think this doesn't account for
me not being allowed to draw myself. This is not a matter of preference,
it's essential:)

-- 
groetjes, carel



From emuller at painewebber.com  Thu Nov 11 18:48:35 1999
From: emuller at painewebber.com (emuller at painewebber.com)
Date: Thu, 11 Nov 1999 23:48:35 GMT
Subject: Win32: CreateProcess wsetup program problem
Message-ID: <80fkki$kmg$1@nnrp1.deja.com>

I wrote a wrapper to CreateProcess that runs a program with args and
waits until it ends. It works great for most programs (ie..calc,
notepad, net, etc, etc, etc), but when I use it to run a setup.exe
program (NAV 5.0 to be exact), the setup program starts and an exit
value of 0 is returned to me, but the setup program just started. I
figure that the setup program is doing something, then spawning another
process to do the install (a _SETUP.EXE..something or other) and
detaching itself from that process, letting it just run on....

Is there anyone out there who knows how I can pause until that other
program (or whatever happens) is done?

I am using the following code in a module called win32helper.py
(BTW: If anyone would like to use this code, go ahead. I hope to write
a bunch of helper functions and put them here, then post them someday)

import win32api
import win32security
import win32process
import pywintypes
import win32con
import win32event

def RunWait(cmd, args=""):
        pHandle, tHandle, pId, tId = win32process.CreateProcess(cmd," "
+ args,None,None,0,0,None,None,win32process.STARTUPINFO())
        win32event.WaitForSingleObject(pHandle,win32event.INFINITE)
        exitCode = win32process.GetExitCodeProcess(pHandle)
        win32api.CloseHandle(pHandle)
        win32api.CloseHandle(tHandle)
        return exitCode


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



From Gareth.McCaughan at pobox.com  Fri Nov 26 17:33:39 1999
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: 26 Nov 1999 22:33:39 +0000
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <867lj85g4c.fsf@g.local>
	<al8yabo9fm3.fsf@sirppi.helsinki.fi> <86so1v4jlj.fsf@g.local>
	<al8hfiaan8p.fsf@sirppi.helsinki.fi> <86ln7m2t31.fsf@g.local>
	<m3emddrmuh.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <86d7sw288s.fsf@g.local>

Michael Hudson wrote:

> This leads to a thought: would it be possible using lisp's silly
> amount of flexibility to get the reader to auto-paranthesize (word?
> doubt it):
> 
> defun our-member (obj lst)
>   if null lst
>      nil
>      if eql (car lst) obj
>         lst
>         our-member obj (car lst)
> 
> based on indentation, as
> 
> (defun our-member (obj lst)
>   (if (null lst)
>     nil
>     (if (eql (car lst) obj)
>       lst
>       (our-member obj (car lst)))))
> 
> or is whitespace ignorance built-in?

It's built in. It might conceivably be possible to hack the
reader to grok whitespace like this, but it would be rather
a pain. It would certainly be possible to write your own
reader (which would call the Lisp reader to do most of its
work) that understood whitespace.

I don't think it would be a good idea, though, any more
than it would be a good idea for Pythonistas moving from C
to pass all their code through a translator that turns
code with {...} into indented code. Or for C programmers
more used to Pascal-like languages to do
  #define BEGIN {
  #define END }
and the like. (I believe the source for the original Bourne
shell is full of that sort of abomination.) It's almost
certainly less effort to learn to live with a language's
syntax than to work around it, and it makes it much less
painful reading other people's code or showing them yours.

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
sig under construction



From I.Clarke at strs.co.uk  Tue Nov 30 09:47:19 1999
From: I.Clarke at strs.co.uk (Ian Clarke)
Date: Tue, 30 Nov 1999 14:47:19 +0000
Subject: Recursion?
References: <38421570.9C3D9AEC@videotron.ca>
Message-ID: <3843E377.58EBFF97@strs.co.uk>

> Can I use recursion in python?

yes

--
Ian Clarke                        Email: I.Clarke at strs.co.uk
Homepage: http://www.gnu.demon.co.uk/
Also see: http://freenet.on.openprojects.net/
"All we see and all we seem is but a dream within a dream"



From gvwilson at nevex.com  Mon Nov 29 11:29:30 1999
From: gvwilson at nevex.com (Greg Wilson)
Date: Mon, 29 Nov 1999 11:29:30 -0500
Subject: performance of tuple-less Python?
Message-ID: <HSx04.8911$cz3.321455@nnrp1.uunet.ca>

Has anyone ever built a version of Python that didn't include tuples, and
looked at the performance of code that consistently used lists instead?  I'm
curious as to whether the use of tuples (which I believe are conceptually
redundant) actually has a significant positive impact on the interpreter's
performance on non-trivial codes (i.e. once caching, instruction ordering,
and all other effects have come into play).
Thanks,
Greg






From piet at cs.uu.nl  Wed Nov 10 04:41:35 1999
From: piet at cs.uu.nl (piet at cs.uu.nl)
Date: 10 Nov 1999 10:41:35 +0100
Subject: __str__ vs. __repr__
References: <199911041325.IAA14001@eric.cnri.reston.va.us> <000001bf275c$02370d20$1d2d153f@tim> <u4glOIWXzGrqvJnI+IlPAQeurz1J@4ax.com>
Message-ID: <uhfiuy980.fsf@cs.uu.nl>

>>>>> Toby Dickenson <htrd90 at zepler.org> (TD) writes:

TD> "Tim Peters" <tim_one at email.msn.com> wrote:
>> Let me back off to what repr and str "should do":
>> 
>> repr(obj) should return a string such that
>> eval(repr(obj)) == obj

TD> I don't understand the motivation for this requirement. Why would
TD> anyone want to pass such a string to eval? If you anticipate the need
TD> for reconstructing the object from a textual representation, then
TD> surely pickle is a better option?

for example, if you would want to write a couple of variables to a file in
the form of python statements: var=value.
-- 
Piet van Oostrum <piet at cs.uu.nl>
URL: http://www.cs.uu.nl/~piet [PGP]
Private email: Piet.van.Oostrum at gironet.nl



From s323140 at student.uq.edu.au  Tue Nov  2 10:18:05 1999
From: s323140 at student.uq.edu.au (Rob Hodges)
Date: 03 Nov 1999 01:18:05 +1000
Subject: Assorted Questions regarding the future of Python...
References: <6D8A17398E28D3119F860090274DD7DB49833A@pces.cadlab.it>
Message-ID: <m3bt9cgbwy.fsf@phaedrus.zen.xom>

Alessandro Bottoni <Alessandro.Bottoni at think3.com> writes:

> 3) Visual Basic has two advantages on Python: it can compile programs (OK: I
> know that we can generate "frozen" Python programs with "Freeze") 

Unless something has changed dramatically since I last used windoze,
VB is not compiled (in the sense of native machine code) any more than
a frozen python program; rather, VB "executables" have a small header
of native code that launches the vbrunxxx.dll (interpreter) to run the
bytecode that makes up the rest of the file.  Which is more or less
the equivalent of writing "#! /usr/bin/env python" at the top of your
python script in UN*X, providing a corresponding .pyc has been
generated.  Or double-clicking a .pyc file in windoze, for that
matter.  So either way, your users have to either have the python
interpreter or the appropriate vb dll somewhere on their system.

Someone please correct me if I'm wrong.  (Though there's no way I'll
ever touch VB <*shudder*>; just that I wouldn't like to be spreading
disinformation.)

-Rob




From l.szyster at ibm.net  Mon Nov 29 13:05:31 1999
From: l.szyster at ibm.net (Laurent Szyster)
Date: Mon, 29 Nov 1999 19:05:31 +0100
Subject: XML to in-memory tree?
References: <tRf%3.8219$cz3.293110@nnrp1.uunet.ca>
Message-ID: <3842C06B.84C431D3@ibm.net>

Greg Wilson wrote:
> 
> Hi.  Has anyone built a layer on top of XMLLIB that will load an
> XML document into an in-memory (similar to the one built by
> explicit calls to HTMLgen)?
> Thanks,
> Greg

if you don't mind using a c module (namely pyexpat, the wrapper
around James Clark famous Expat), there is Greg Stein qp_xml.py

you can find it somewhere at

  <http://www.lyra.org/greg/python/>

it's not very orthodox, but quite convenient


Laurent Szyster



From emile at fenx.com  Thu Nov 25 11:01:16 1999
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 25 Nov 1999 08:01:16 -0800
Subject: Alphabetized dictionary listings
Message-ID: <053501bf375e$97098420$01ffffc0@worldnet.att.net>

Ken,

It sounds like you're looking for something like this.


#-------Start Code Fragment--------
def display(d):
 keys = d.keys()
 keys.sort()
 for key in keys:
  print "%3d %10s : %s" % (d[key][1], key, d[key][0])

def putkeys(mDict, mList):
 " put keys only in a dict while preserving entry sequence "
 seq = len(mDict)
 for key in mList:
  mDict[key]=[[],seq]
  seq = seq + 1

def putdata(mDict, mList):
 " put data in dict by preserved sequence "
 for key, data in mDict.items():
  data[0]=mList[data[1]]
  mDict[key] = data

myDict = {}
myKeys = ['Abel', 'Dexter', 'Francis', 'Matthew', 'Samantha', 'Tyrone']
myData = ['lebA', 'rexteD', 'sicnarF', 'wehttaM', 'ahtnamaS', 'enoryT']

putkeys(myDict, myKeys)
putdata(myDict, myData)

myNewKeys = ['AAbel', 'DDexter', 'FFrancis', 'MMatthew', 'SSamantha',
'TTyrone']
myCompleteData = ['lebA', 'rexteD', 'sicnarF', 'wehttaM', 'ahtnamaS',
'enoryT', 'lebAA', 'rexteDD', 'sicnarFF', 'wehttaMM', 'ahtnamaSS',
'enoryTT']

putkeys(myDict, myNewKeys)
putdata(myDict, myCompleteData)

display(myDict)
#-------End Code Fragment--------

--

Emile van Sebille
emile at fenx.com
-------------------


Ken Power <iam at not.you> wrote in message
news:383529e0.682634 at news1.mysolution.com...
> After reading my post I realized I undervalued what I am really
> attempting to accomplish. Thank you everyone for there help to this
> point. The info about hash values now makes sense. Ok, here is what I
> am really trying to do:
>
> Give a value to a key from a list. However, the values in a list are
> in a prescribed order and must be linked to a specific key in the
> dictionary. I tried to demonstrate that in the previous post, linking
> the list values (names spelled backward) to a specific key in the
> dictionary. In my example, I wanted the names to match. Understandably
> if I placed the values in a dictionary in the first place, I could use
> the keys to match each other, however that would be ridiculous :),
> especially since my goal is 'dynamicly' created dictionaries
> (dictionaries created during run-time, I guess).  O well, I ramble
> now.
> Here is what I wanted as output from the previous example:
> >>>> for key in myDict.keys():
> > print key, myDict[key]
> >
> >
> >Tyrone enoryT
> >Samantha  ahtnamaS
> >Francis  sicnarF
> >Abel  lebA
> >Matthew  wehttaM
> >Dexter  retxeD
>
> Guess I'll have to approach the problem from another angle......
> --------------------------------
> Ken Power
> uncle_wiggly at bigfoot dot com
> --------------------------------
>
> --
> http://www.python.org/mailman/listinfo/python-list
>
>





From cjensen at be-research.ucsd.edu  Wed Nov 17 14:54:30 1999
From: cjensen at be-research.ucsd.edu (Curtis Jensen)
Date: Wed, 17 Nov 1999 11:54:30 -0800
Subject: Multi-Dimensional Arrays.
References: <3832F2B9.F64C8AD7@be-research.ucsd.edu>
Message-ID: <383307F6.B8F02BD7@be-research.ucsd.edu>

Allow me to refine the question.  If I make a list of lists to be my
multi-dimensional array, how do I pass that into C?
PyArg_ParseTuble(args, "?", &x)?  What would go in place of the "?"?  Or
would it be done a different way.  Then to pass it back to Python, I
would return what?  Could I use Py_BuildVal("?", x) ?

Curtis Jensen wrote:

> I'm working on a Python user interface that will be connected to some
> fortran code with heaps of multi-dimensional arrays.  I can't seem to
> find any references to multi-dimensional arrays in Python.  Is there
> an easy way to create multi-dimensional arrays in Python?
>
> --
> Curtis Jensen
> cjensen at be-research.ucsd.edu
> http://www-bioeng.ucsd.edu/~cjensen/
> FAX (425) 740-1451
>
>

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


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/19991117/ce1dc4ed/attachment.html>

From jeremy at cnri.reston.va.us  Tue Nov  2 11:51:54 1999
From: jeremy at cnri.reston.va.us (Jeremy Hylton)
Date: Tue, 2 Nov 1999 11:51:54 -0500 (EST)
Subject: newbie intro recs?
In-Reply-To: <381E3205.BEFDE12C@home.com>
References: <381E3205.BEFDE12C@home.com>
Message-ID: <14367.5802.984523.523601@goon.cnri.reston.va.us>

It sounds like Learning Python would be perfect for you.  It assumes
you have programmed before, but not that you are expert in one or many
languages.

Jeremy




From culliton at clark.net  Tue Nov  2 09:51:31 1999
From: culliton at clark.net (Tom Culliton)
Date: Tue, 02 Nov 1999 14:51:31 GMT
Subject: (off topic) Re: About RedHat 6.1 install program
References: <7vm92d$i1e$1@nnrp1.deja.com> <000701bf2514$01458490$f29b12c2@secret.pythonware.com>
Message-ID: <TTCT3.32291$oa2.44407@iad-read.news.verio.net>

In article <000701bf2514$01458490$f29b12c2 at secret.pythonware.com>,
Fredrik Lundh <fredrik at pythonware.com> wrote:
>johndoe0521 at my-deja.com wrote:
>> I think this might be python's problem.
>
>have you contacted redhat?  did they tell you
>to ask here?
>
>if not, try:
>
>http://www.redhat.com/cgi-bin/support_switch
>
>if they *did* tell you to ask here, you most provide
>more information.  what did they say?  what errors

There are at least two problems, first a mount/unmount problem in Red
Hat's logic, second an extra comma on line 1508 line of todo.py.  The
real trigger seems to be when you run out of disk space after packaage
selection and have to back up.  Go to the following URL, then to the
query page, and query for the "Red Hat Linux", "6.1", "installer"
package, and you should find several (178) bugzilla items possibly
related to your problem.  Including "python" in the description field
as a criteria will narrow the list some (31 items).  ;-)

http://developer.redhat.com/bugzilla/

Bottom line the new Anaconda installer is NEW, and like too mch new
software hasn't had enough of the wrinkles shaken out.  The switch to
gnorpm in RH 6.0 was just as bad, the app was several steps back from
glint which it replaced, the version with 6.1 is miles better but I
still know at least 3 ways to crash it.  Over all Red Hat tends to
push stuff out the door a little early for my taste.



From python-list at teleo.net  Sat Nov 13 11:47:37 1999
From: python-list at teleo.net (Patrick Phalen)
Date: Sat, 13 Nov 1999 08:47:37 -0800
Subject: A Modest Proposal -- <wink>
In-Reply-To: <y0j9042tlrx.fsf@vier.idi.ntnu.no>
References: <99110613580407.02970@quadra.teleo.net> <80jqqo$n3m$1@newshost.accu.uu.nl> <y0j9042tlrx.fsf@vier.idi.ntnu.no>
Message-ID: <99111308583701.02728@quadra.teleo.net>

[Magnus L. Hetland, on Sat, 13 Nov 1999]:

:: But - if you have XPPMS, why do you need XWTSNTSDL? You could just
:: transform
:: 
::    This is some text <wink>
:: 
:: into
:: 
::    <wink>This is some text</wink>
:: 
:: and Bob's your uncle... Then you can use standard XML tools... :)


I am against this notion, for several reasons. 

1. XPPMS should disavow any attempt to do anything useful. I thought we
had all agreed to this. It's in the spec!

2. It gratuitously introduces containment and well-formedness, which,
in turn, threatens several key objectives of XWTSNTSDL, XGNBAML, and
XACL, namely:

* Make XML software significantly harder to write
* Make source files significantly harder to read

We must be especially cautious not to introduce any changes at
this point which might make XML easier to teach to beginners.




From python-list at teleo.net  Sat Nov  6 15:11:43 1999
From: python-list at teleo.net (Patrick Phalen)
Date: Sat, 6 Nov 1999 12:11:43 -0800
Subject: JPython and NetBeans and Java envy (Was Re: Python and Smalltalk and Java envy)
In-Reply-To: <382484F1.F8B8D712@home.com>
References: <1270663667-24391282@hypernet.com> <7vqkqq$oa7$1@nnrp1.deja.com> <382484F1.F8B8D712@home.com>
Message-ID: <99110612201506.02970@quadra.teleo.net>

[Wostenberg, on Sat, 06 Nov 1999]:

:: Intersting observation. Perhaps it comes down to those of us who are
:: comfortable with niche languages (Python, Smalltalk, Lisp) and those who
:: suffer Java-envy . Do Python people, Ada advocates, or little Lispers wring
:: there hands about the fact "we are not the most used language?" Why should
:: Smalltalkers?


Speaking of Java envy, Barry Warsaw pointed out the other day on the
JPython list that www.netbeans.com is running a survey to pick the best
scripting language to be integrated with NetBeans. (You'll find the form
in their left navbar.)

I won't reveal the *overwhelming* winner, thus far. Find out for
yourself.  ;)

I-swear-I-only-voted-twice-ly-y'rs   -patrick



From gawron at obop.com.pl  Fri Nov 19 06:16:22 1999
From: gawron at obop.com.pl (=?iso-8859-2?Q?Przemys=B3aw?= G. =?iso-8859-2?Q?Gawro=F1ski?=)
Date: Fri, 19 Nov 1999 12:16:22 +0100
Subject: Running Exel
Message-ID: <38353186.9D9124EF@obop.com.pl>

How can I run MS Exel via OLE using Python (with windows ext.) ?

Thankx

Przemek

--
Przemyslaw G. Gawronski UIN:8358522
mailto:gawronskip at usa.net
mailto:gawron at obop.com.pl





From tiddlerdeja at my-deja.com  Mon Nov  8 12:40:11 1999
From: tiddlerdeja at my-deja.com (tiddlerdeja at my-deja.com)
Date: Mon, 08 Nov 1999 17:40:11 GMT
Subject: Psion JVM
Message-ID: <8071tr$9ar$1@nnrp1.deja.com>

Has anyone here tried JPython on a Psion JVM (like the 5mx I think).

Any information is appreciated.


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



From MHammond at skippinet.com.au  Thu Nov  4 18:14:52 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Thu, 04 Nov 1999 23:14:52 GMT
Subject: return boolean from functions
References: <7vrq9b$hmr$1@nnrp1.deja.com> <m3iu3ia1th.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <MroU3.868$pf4.4710@news-server.bigpond.net.au>

Michael Hudson wrote in message ...
>a_olme at my-deja.com writes:
>
>Well, you could return 1 for true and 0 for false:
>
>>>> print 1==1, type(1==1), 1==0, type(1==0)
>1 <type 'int'> 0 <type 'int'>
>
>Python does not have a distinct boolean type, if that's what you're
>asking. Empty lists, empty strings, 0, 0L and 0.0 are false,
>everything else is true (except for instances of classes that have
>__nozero__ methods, which get to decide for themselves).

Actually, Python sort-a does have a concept of booleans.  There are PyTrue
and PyFalse objects, but they are not exposed to the runtime in any
meaningful way.  Eg:

>>> 1==1
1
>>> 1 is 1
1
>>> (1==1) is 1
0
>>>

Not that this really means anything practical, but there you have it...

Mark.





From sonic at sega.net  Tue Nov 30 16:47:14 1999
From: sonic at sega.net (sonic at sega.net)
Date: Tue, 30 Nov 1999 21:47:14 GMT
Subject: Grabbit two now available for FREE  6200
Message-ID: <CBX04.73892$up3.113481@news21.bellglobal.com>

This amazing little has now got a whole lot better, and Costech is please to offer it to our friends for free.  Enjoy the power 

http://www.costech.com/Products/Grabit/grabit.html
vwtrzsvweev




From lobozc at my-deja.com  Wed Nov 24 17:17:06 1999
From: lobozc at my-deja.com (lobozc at my-deja.com)
Date: Wed, 24 Nov 1999 22:17:06 GMT
Subject: filter function question - list processing in Python
References: <81f18b$hor$1@nnrp1.deja.com> <383BF71F.3E413E33@compaq.com>
Message-ID: <81ho50$g15$1@nnrp1.deja.com>

1. Thank you for your answers. I infer from them that my reading of
_standard_ list processing functions in Python was correct.

2. I suspect that the second part of my posting may be a valid
postulate then. Since list processing is so useful Python should
consider adding a good set of primitives to the core library - not as
an external package.

In article <383BF71F.3E413E33 at compaq.com>,
  Greg Ewing <greg.ewing at compaq.com> wrote:
> lobozc at my-deja.com wrote:
> >
> > How should I pass additional parameter in 'filter'? 'map' has them,
> > filter doesn't - why? Say I have a list ['xxx.py','xxx.pyc'] and I
want
> > to have a function isExtension(x,'.py')
> > [No, I'm not interested in using lambda here in general].
>
> Why not? That's the shortest way to do it:
>
>    def filterExt(items, ext):
>       return filter(lambda x, ext=ext: isExtension(x,ext), items)
>
> However, a more Pythonic way of writing that is
>
>    def filterExt(items, ext):
>       result = []
>       for x in items:
>          if isExtension(x, ext):
>             result.append(x)
>       return result
>
> which is longer-winded but clearer. If you want something which
> is both short-winded and clear, you might like to try my List
> Comprehension syntax extension, which allows you to write
>
>    def filterExt(items, ext): # NOT STANDARD PYTHON
>       return [x for x in items if isExtension(x, ext)]
>
> The List Comprehensions patch is available from my web page:
> http://www.cosc.canterbury.ac.nz/~greg
>
> Greg
>
>  - which I can call with various
> > extensions to filter the list in variuos ways - how am I supposed
to do
> > that? [No, I'm not interested in using lambda here in general].
> >
> > In general, I note that Python has some standard functions
operating on
> > lists but they are curiously limited. It is a pity, since these are
the
> > functions which reduce line count significantly - as well as make
> > programming safer. I wonder what are the chances of repeating with
> > lists what Python did with regular expressions - copying the
> > functionality from a well tested and popular product?
> >
> > Mathematica has a set of list processing functions. As one can
expect
> > from an essentially functional language and a product where most
> > operations are done by list processing - for the last 10 years -
this
> > set is well thought out, complete and very, very useful.
> >
> > [For description see
> > http://documents.wolfram.com/v4/frames/frames.html, press
button 'built-
> > in functions', on the leftmost panel 'programming', to the right of
it
> > press 'functional programming' - eventually look for functions like
> > Map, Fold, Apply etc.].
> >
> > Sent via Deja.com http://www.deja.com/
> > Before you buy.
>


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



From tellis at giasbma.vsnl.net.in  Wed Nov 24 11:58:51 1999
From: tellis at giasbma.vsnl.net.in (Philip Tellis)
Date: Wed, 24 Nov 1999 16:58:51 +0000
Subject: Absolute newbie question
Message-ID: <Pine.SOL.3.96.991124165729.11616B-100000@giasbma>

Where are the man pages for python?  I am new to python, and would like to
look through the man pages before I start.  Someone please help.

Philip





From gmcm at hypernet.com  Tue Nov  9 09:58:02 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 9 Nov 1999 09:58:02 -0500
Subject: README: Python-URL! - the weekly guide to Python resources (Nov 9)
Message-ID: <1269963015-19149073@hypernet.com>

 This week on comp.lang.python we find:

 Evan Simpson's one page description of the anatomy of Zope:
   http://www.deja.com/getdoc.xp?AN=544506548

 Paul Duffin's speculation about how "stubbed" Tcl might be
   used by Tkinter:
   http://www.deja.com/getdoc.xp?AN=545858112

 The __str__ vs __repr__ thread rages on.
   Guido explains the rationale:     
      http://www.deja.com/getdoc.xp?AN=544209311
   Tim clarifies the problem:
      http://www.deja.com/getdoc.xp?AN=544718491
   Bernhard Herzog makes an interesting proposal:      
     http://www.deja.com/getdoc.xp?AN=544360351
   while others split on whether "eval(repr(obj)) == obj" is a
   valuable property or a wart.

 Jean-Claude Wippler announces a stable beta of his Python
   wrappers for MetaKit (a cross-platform embedded database
   with intriguing properties):
   http://www.equi4.com/metakit/python.html

 Regsitration for IPC8 is now online:
   http://www.python.org/workshops/2000-01/registration.html
   and prompts some discussion of travel arrangements.

 Two new Python books get rave reviews:
   David Beazley's Python Essential Reference
   and Fredrik Lundh's ebook:
    (the eff-bot guide to) The Standard Python Library
   (although technical / platform problems create some
    controversy for the eff-bot!).
   http://www.python.org/psa/bookstore/

 Greg Ewing compares Python and Ruby and makes some
   interesting observations: 
    on extension writing interface:
     http://www.deja.com/getdoc.xp?AN=543792249
    on the object model:
     http://www.deja.com/getdoc.xp?AN=544255921

 Python scores points on rec.humor.funny:
     http://www.netfunny.com/rhf/jokes/99/Nov/perl.html

=============================================

Everything you want is probably one or two clicks away in 
these pages:

      Python.org's Python Language Website is the center of Pythonia
          http://www.python.org

      Much of Python's real work takes place on Special-Interest Group
          mailing lists http://www.python.org/sigs/

      Python Consortium emerges as an independent nexus of activity
          http://www.python.org/consortium

      Python To-Do List anticipates some of Python's future direction
          http://www.python.org/cgi-bin/todo.py

      Python Journal is at work on its second issue
          http://www.pythonjournal.com

      Links2Go is a new semi-automated link collection; it's impressive
      what AI can generate
          http://www.links2go.com/search?search=python

    Archive probing trick of the trade:

http://www.dejanews.com/dnquery.xp?QRY=&DBS=2&ST=PS
&defaultOp=AND&LNG=AL

L&format=threaded&showsort=date&maxhits=100&groups=co
mp.lang.python


Previous - (U)se the (R)esource, (L)uke! - messages are listed 
here:
  http://purl.org/thecliff/python/url.html
or

http://www.dejanews.com/dnquery.xp?QRY=~g%20comp.lang.
python%20Python-URL%21


Suggestions/corrections for next week's posting are always 
welcome.
  http://www.egroups.com/list/python-url-leads/

To receive a new issue of this posting in e-mail each Monday 
morning,
ask <claird at neosoft.com> to subscribe.  Mention "Python-
URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to 
participate in and
sponsor the "Python-URL!" project.
-- 

Cameron Laird           
http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX




From lobozc at my-deja.com  Tue Nov 23 16:34:10 1999
From: lobozc at my-deja.com (lobozc at my-deja.com)
Date: Tue, 23 Nov 1999 21:34:10 GMT
Subject: filter function question - list processing in Python
Message-ID: <81f18b$hor$1@nnrp1.deja.com>

How should I pass additional parameter in 'filter'? 'map' has them,
filter doesn't - why? Say I have a list ['xxx.py','xxx.pyc'] and I want
to have a function isExtension(x,'.py') - which I can call with various
extensions to filter the list in variuos ways - how am I supposed to do
that? [No, I'm not interested in using lambda here in general].

In general, I note that Python has some standard functions operating on
lists but they are curiously limited. It is a pity, since these are the
functions which reduce line count significantly - as well as make
programming safer. I wonder what are the chances of repeating with
lists what Python did with regular expressions - copying the
functionality from a well tested and popular product?

Mathematica has a set of list processing functions. As one can expect
from an essentially functional language and a product where most
operations are done by list processing - for the last 10 years - this
set is well thought out, complete and very, very useful.

[For description see
http://documents.wolfram.com/v4/frames/frames.html, press button 'built-
in functions', on the leftmost panel 'programming', to the right of it
press 'functional programming' - eventually look for functions like
Map, Fold, Apply etc.].




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



From edwardam at home.com  Tue Nov 30 00:01:59 1999
From: edwardam at home.com (Edward Muller)
Date: Tue, 30 Nov 1999 05:01:59 GMT
Subject: 50% Python content :)   (Distributed computing)
References: <816010E2456BD111A48700805FBBE2EEFDF862@ex-quebec-u1.baan.com>
Message-ID: <38435A43.34A434BE@home.com>

It's really a matter of what you want to do. So, what do you want to do?

I've used xmlrpclib (see www.pythonware.com for the Python XML-RPC Library; see
www.xml-rpc.com for details on the specs) very successfully in the field and
for personal use. It's a great way to conact another PC, have it do something
and return the results.


Gaetan Corneau wrote:

> Hello,
>
> Sorry if my question is not strictly a Python question.
>
> I would like to explore distributed programming, and I think Python would be
> the right tool for that. I don't want to produce industrial-strenght code
> right now, but learn the basics. I'm not really interested in the
> "middleware" aspects, so I didn't subscribe to the "Python Distributed
> Objects" SIG right away. I just want to answer questions like how do I
> partition a monolithic program to run on multiple computers?
>
> I went to the bookstore, and found some title (not with Python), but I
> wasn't sure which one to buy. Those are expensive, and I thought that
> someone here could point me in the right direction. I'm looking for
> practical, hands-on stuff, if possible.
>
> Thanks in advance,
> ______________________________________________________
>    Gaetan Corneau
>    Software Developer (Quality Assurance Team)
>    BaaN  Supply Chain Solutions
>    http://www.baan.com
>    E-mail: Gaetan_Corneau at baan.com
>    Tel: (418) 266-8252
> ______________________________________________________
> "Profanity is the one language all programmers know best"




From I.Clarke at strs.co.uk  Thu Nov 11 07:01:26 1999
From: I.Clarke at strs.co.uk (Ian Clarke)
Date: Thu, 11 Nov 1999 12:01:26 +0000
Subject: Intelligent strptime?
References: <19991110171256.A5196@optiplex.palga.uucp>
Message-ID: <382AB016.D45727D8@strs.co.uk>

> Does there exist an intelligent strptime? "26/11/99", for example, can _only_
> mean the 26th November of this year, and the chance that "14:11:12" means
> 14 november 1912 is very small. Is there a module/function that guesses the
> format? There isn't doubt very often, only for 12/11/99, US is 11 december,
> Europe is 12 November (isn't it?).

While I don't know what the application is, if you are processing user
input I strongly recommend against doing this.  In a project I am
working on the issue of whether to provide a "best guess" as to a date
came up, but we decided that it was much better if a date had been
entered incorrectly to furnish the user with an error, than attempt to
correct that error automatically.  Sure in the short term it makes life
a little easier for the user, but it also allows them to get used to the
computer correcting their errors, and when the time comes that the
computer cannot correct their error, or makes an incorrect guess, they
will feel that there is something wrong ("I have always entered dates in
MMDDYY format and it worked fine, why shouldn't it work now?").

Ian.

--
Ian Clarke                        Email: I.Clarke at strs.co.uk
Homepage: http://www.gnu.demon.co.uk/
Also see: http://freenet.on.openprojects.net/
"All we see and all we seem is but a dream within a dream"



From loewis at informatik.hu-berlin.de  Mon Nov 29 14:45:28 1999
From: loewis at informatik.hu-berlin.de (Martin von Loewis)
Date: 29 Nov 1999 20:45:28 +0100
Subject: pre/post-condition support?
References: <_Re04.8705$cz3.315483@nnrp1.uunet.ca>
Message-ID: <p6q1z99m693.fsf@informatik.hu-berlin.de>

"Greg Wilson" <gvwilson at nevex.com> writes:

> Has anyone experimented with Python extensions to support programming by
> contract (pre/post conditions on methods + state assertions that are
> respected by inheritance)?  If so, I'd welcome pointers.

As for pre/post conditions and state assertions: with Python 1.5.x,
there is the assert statement:

class Data:
  def __init__(self):self.items = {}
  def add(self,key,value):
    assert (not self.items.has_key(key))
    self.items[key] = value

d = Data()
d.add("milk","white")
d.add("milk","white")

As for respecting inheritance: I guess approaches taken in other
languages work in Python as well:

class Base:
  def method(self,params):
    assert preconditions
    result = self.method_implementation(self,params)
    assert postconditions
    return result

You can probably apply a number of reflection-based tricks in Python;
I doubt that this would make your code more understandable.

Regards,
Martin



From lobozc at my-deja.com  Thu Nov 25 16:04:20 1999
From: lobozc at my-deja.com (lobozc at my-deja.com)
Date: Thu, 25 Nov 1999 21:04:20 GMT
Subject: filter function question - list processing in Python
References: <1268631776-5935125@hypernet.com>
Message-ID: <81k88j$5m3$1@nnrp1.deja.com>

> all-functional-programmers-are-rabid-ly y'rs
And where's the smile at the end of that sentence? :-)

> But, I suspect, those who regularly use filter and reduce are
> vastly outnumbered by those who aren't sure what they do. So
yeah, only now I noticed that examples of their use are conspicuous by
their absence...

... and I don't have time to write the package right now, either. Nor
enough python skills.

In article <1268631776-5935125 at hypernet.com>,
  gmcm at hypernet.com wrote:
> lobozc at my-deja.com writes:
>
> > 2. I suspect that the second part of my posting may be a valid
> > postulate then. Since list processing is so useful Python should
> > consider adding a good set of primitives to the core library -
> > not as an external package.
>
> A good set of primitives with intuitive interfaces and obvious
> names that raises lots of excitement would probably be
> warmly welcomed.
>
> But, I suspect, those who regularly use filter and reduce are
> vastly outnumbered by those who aren't sure what they do. So
> rabid enthusiasm may be hard to come by.
>
> all-functional-programmers-are-rabid-ly y'rs
>
> - Gordon
>
>


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



From steve at acer.spvi.com  Sun Nov  7 22:28:21 1999
From: steve at acer.spvi.com (Steve Spicklemire)
Date: Sun, 7 Nov 1999 22:28:21 -0500 (EST)
Subject: Python for children.....
In-Reply-To: <801p16$10s$1@vvs.superst.iae.nl> (message from Carel Fellinger
	on 6 Nov 1999 18:37:42 +0100)
References: <9911051740130D.02530@william.jinsa.org> <7GOU3.272$Ae3.194028@news-dal.corridex.com> <801p16$10s$1@vvs.superst.iae.nl>
Message-ID: <199911080328.WAA24200@acer.spvi.com>

I've also been playing with teaching python to 'children' (well..
middle schoolers anyway... don't tell *them* they are children. ;->)
I've had some luck with my little PyBots toy...  it needs some work,
when I get more time.. but it's enough to teach some simple stuff...

Basically it's a Macromedia Director movie with a single 'cast
member', a circle with a tab. On the mac version the movie imports a
python module (called robots.py) and calls three fuctions over and
over until the user presses the quit button. The functions are:
getX(), getY() and getRot(). They return the horizontal position,
vertical position, and angle of rotation of the ball. With just that
you can teach them if, while, for, functions, object orientation etc..
for a slightly advanced example I came up with this:

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

import random

class MovingObject:

    x = 100
    y = 100
    vx = 4
    vy = 4
    omega = 1
    theta = 0

    def getX(self):

        if self.x > 300:
            self.vx = -4 + 2*random.random()
        if self.x < 40:
            self.vx = 4 + 2*random.random()

        self.x = self.x + self.vx
        return self.x


    def getY(self):

        if self.y > 300:
            self.vy = -4 + 2*random.random()
        if self.y < 40:
            self.vy = 4 + 2*random.random()

        self.y = self.y + self.vy

        return self.y

    def getRot(self):
        if self.theta > 360:
            self.omega = -1  + 2*random.random()
        if self.theta < 0:
            self.omega = 1 + 2*random.random()

        self.theta = self.theta + self.omega
        return self.theta

m = MovingObject()

def getX():
    return m.getX()

def getY():
    return m.getY()

def getRot():
    return m.getRot()

if __name__=="__main__":
    print getX()
    print getY()
    print getRot()

                                
Evan Simpson <evan at tokenexchange.com> wrote:

> Now I'm preparing to introduce my children to programming through Python.
> I'm starting with their fascination with Alice, the 3-D animation workshop
> which even my 6 year old enjoys playing with (though he doesn't script it...
> yet).






From ivanlan at callware.com  Tue Nov  2 16:51:13 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Tue, 02 Nov 1999 14:51:13 -0700
Subject: Reading .lnk files
Message-ID: <381F5CD1.8714622F@callware.com>

Hi All--
Is there a function in the Win32 stuff to read .lnk files?  I'm looking
for a *minimal* script to do nothing but read a .lnk given on the
command line and spit the contents out on stdout.

Thanks,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From news at dorb.com  Wed Nov 10 16:16:14 1999
From: news at dorb.com (Darrell)
Date: Wed, 10 Nov 1999 16:16:14 -0500
Subject: random function
References: <3829D78B.7FC142F4@scripps.edu>
Message-ID: <ZelW3.5541$JD1.365515@typhoon.nyroc.rr.com>

Python Library Reference
random ()
Returns the next random floating point number in the range [0.0 ... 1.0).

>>> import random
>>> random.random()
0.981053289833
>>>

--
--Darrell
Olivier Roche <olivier at scripps.edu> wrote in message
news:3829D78B.7FC142F4 at scripps.edu...
> Hi everybody
>
> By looking at the index of the python documentation index i haven't seen
> any
> random function description. Could somebody tell me if a random function
>
> to generate random number (between 0 and 1) exists ?
>
> Thanks,
>
> Olivier
>
>
> --
> http://www.python.org/mailman/listinfo/python-list





From timmuddletin at news.vex.net  Sun Nov 28 16:32:08 1999
From: timmuddletin at news.vex.net (tim muddletin)
Date: 28 Nov 1999 21:32:08 GMT
Subject: "Import error"--MySQLmodule broke betw. Python 1.5.1 and 1.5.2
References: <38418F71.A39C213E@dds.nl>
Message-ID: <81s70o$1nne$1@hub.org>

On Sun, 28 Nov 1999 21:24:17 +0100, Martian <martian1 at dds.nl> wrote:
>The MySQLmodule hasn't changed--indeed, hasn't since Oct. 1998

Is there any reason you need to use the old module? According to my
files ( http://www.vex.net/~x/parnassus/apyllo.py?i=82701837 ), there
seems to be a newer (and quite improved sounding) MySQL module here...

 http://starship.python.net/crew/adustman/MySQLdb.html

(note: i don't know anything about it, never having used MySQL.
I'm just a humble python resource librarian. <-: Though i don't
seem to have record of the old module in my database! Hmmm...)






From skip at mojam.com  Fri Nov  5 14:23:37 1999
From: skip at mojam.com (Skip Montanaro)
Date: Fri, 5 Nov 1999 13:23:37 -0600 (CST)
Subject: Can Zope be used as a CVS server?
In-Reply-To: <harms-0511991330570001@192.168.1.35>
References: <harms-0511991330570001@192.168.1.35>
Message-ID: <14371.11961.82152.303247@dolphin.mojam.com>

    Stefan> Could I use Zope for the functionality of a CVS server?

Dunno, and it's not obvious (to me) why you'd want to do that.  Could you
motivate your question by providing some rationale for not using CVS itself?

Skip Montanaro | http://www.mojam.com/
skip at mojam.com | http://www.musi-cal.com/
847-971-7098   | Python: Programming the way Guido indented...



From aa8vb at yahoo.com  Mon Nov 22 13:46:10 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Mon, 22 Nov 1999 13:46:10 -0500
Subject: Python mymalloc.h bug?  (was: Re: Building wxPython)
In-Reply-To: <00e301bf32c6$e77aee90$0301a8c0@easystreet.com>
References: <80s0b4$g5n@mail.psy.uva.nl> <19991117090214.B807152@vislab.epa.gov> <017a01bf31f8$6bbea590$1a25d2d1@jenkondev.com> <19991119064944.A996886@vislab.epa.gov> <002301bf32b5$97a5d440$0301a8c0@easystreet.com> <19991119134422.A1004588@vislab.epa.gov> <00e301bf32c6$e77aee90$0301a8c0@easystreet.com>
Message-ID: <19991122134610.A1255370@vislab.epa.gov>

Robin Dunn:
 |>NULL is defined as ((void*)0) and cannot be assigned to / compared with
 |>another type without a cast in C++.  As this is C++, I believe they
 |>should be using "0" for assignments and comparisons.
 |
 |I don't know how accurate this is but I've been told that C++ compilers that
 |have NULL defined as ((void*)0) in their headers are the ones that are
 |incorrect, not the code that uses NULL.  The headers are supposed to define
 |it as 0.  Maybe you could try undef'ing it and redefining NULL to be 0 and
 |see what happens?

I think you're right.  However the IRIX headers are correct:

     /usr/include/stdio.h:#define NULL               0L

After a bit of grepping, the culprit seems to be Python (1.5.2):

     /usr/local/include/python1.5/mymalloc.h:

         #ifndef NULL
         #define NULL ((ANY *)0)
         #endif

Including <stdio.h> at the top of helpers.cpp before you include anything
else will work around the problem for that source file, but of course
that's a hack, not a fix.

It appears to me that the right fix is for Python's mymalloc.h, if it must
define NULL (does it?), be updated to be more C++ friendly.  E.g.:

         #ifndef NULL
         #  ifdef __cplusplus
         #    define NULL 0
         #  else
         #    define NULL ((ANY *)0)
         #  endif
         #endif

I'm assuming there's some reason we can't just #include <stdio.h> from 
mymalloc.h (?)

-- 
Randall Hopper
aa8vb at yahoo.com



From kuncej at mail.conservation.state.mo.us  Tue Nov  2 14:46:31 1999
From: kuncej at mail.conservation.state.mo.us (Jeffrey Kunce)
Date: Tue, 02 Nov 1999 13:46:31 -0600
Subject: Download Python Imaging Library Handbook
Message-ID: <s81eeb86.053@mail.conservation.state.mo.us>

I had an old version of the handbook, and just got the new one today.
I used websucker.py from the python Tools/webchecker directory.

 >python websucker.py http://www.pythonware.com/library/pil/handbook/index.htm

  --Jeff




From arcege at shore.net  Mon Nov  8 14:04:58 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Mon, 08 Nov 1999 19:04:58 GMT
Subject: debuging in shared library (using gdb or ddd)
References: <3821F93B.6BF2B62A@iwr.uni-heidelberg.de> <Pine.LNX.4.10.9911041645260.9670-100000@us2.mayo.edu>
Message-ID: <u9FV3.397$CT1.80511@news.shore.net>

> Hi!
> 
> How can I debug a shared library module imported into python
> 
> with gdb or ddd debugger ?

I often start gdb with python, run the program getting an interactive
shell, import the appropriate module(s) or package(s).  Then I interrupt
the process (cntl-C), which is captured first by GDB, not Python, set
my breakpoints and finally "cont"inue the interactive interpreter:

$ gdb python
...
(gdb) run
...
>>> import foobar
>>> <cntl-c>
(gdb) break foobar_function
(gdb) cont
Continuing.
<return>
>>>

There may be some configurations where this won't work (threads, etc.),
but I've used it on a lot of systems this way.

  -Arcege




From evan at tokenexchange.com  Tue Nov 16 10:45:39 1999
From: evan at tokenexchange.com (Evan Simpson)
Date: Tue, 16 Nov 1999 09:45:39 -0600
Subject: Zope comparisons? (ASP, PHP, etc)
References: <3821C5F6.84543AF2@yahoo.com><06lU3.259$Ae3.186544@news-dal.corridex.com> <B455938E.5FE6%petrilli@digicool.com>
Message-ID: <Y%eY3.384$Ae3.251645@news-dal.corridex.com>

Christopher Petrilli <petrilli at digicool.com> wrote in message
news:B455938E.5FE6%petrilli at digicool.com...
> in article 06lU3.259$Ae3.186544 at news-dal.corridex.com, Evan Simpson at
> evan at tokenexchange.com wrote on 11/4/99 2:26 PM:
> > single file (2GB limit being worked on) and appends new versions of
objects
>
> Just want to clarify that this is a Linux restriction, not a Zope
> restriction.  We use a 64-bit OID, so if the underlying OS supports larger
> files, then we should continue to grow.

Sorry, I should have phrased that better, and would have if I had known how
widely read this posting would be :-)

How about "(the capability to mount multiple storages in a single tree is
being worked on, which will circumvent the Linux 2GB file-size limitation
and provide other advantages)".

Cheers,

Evan @ 4-am





From amitp at Xenon.Stanford.EDU  Tue Nov  2 21:48:08 1999
From: amitp at Xenon.Stanford.EDU (Amit Patel)
Date: 3 Nov 1999 02:48:08 GMT
Subject: Default curses module.
References: <19991031170934.A18715@optiplex.palga.uucp>
Message-ID: <7vo7p8$d9a$1@nntp.Stanford.EDU>

 Gerrit Holl  <gerrit at nl.linux.org> wrote:
| Hello,
| 
| why isn't the newest cursesmodule, available at
| http://andrich.net/python/selfmade.html#ncursesmodule, in the standard python
| distribution?
| 
| I don't want all my users to have to recompile python for using my
| colors-using program!

I have avoided using colors, because I don't want my users to have to
compile this module.  :-(

The explanation I've heard is that ncursesmodule isn't as portable as
cursesmodule, so it didn't go into the standard ..

	      - Amit



From phd at phd.russ.ru  Fri Nov 12 08:29:15 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Fri, 12 Nov 1999 13:29:15 +0000 (GMT)
Subject: time/strptime/mktime/localtime
In-Reply-To: <382C0B8B.DE2DC3A9@inka.de>
Message-ID: <Pine.LNX.4.20.9911121328180.21950-100000@fep132.fep.ru>

On Fri, 12 Nov 1999, Michael [iso-8859-1] Str?der wrote:
> > 
> > print time.strptime("1999-09-21 21:44:33", "%Y-%m-%d %T")
> > 
> > Linux:   (1999, 9, 21, 21, 44, 33, 6, 1, 0)
> > Solaris: (1999, 9, 21, 21, 44, 33, 6, 1, 0)
> 
> A couple of weeks ago I found some oddities with strptime on my
> glibc-based Linux system. Someone told me that strptime in glibc is
> broken.

   Other way around - I have problems on Solaris. On linux all goes well.

Oleg.
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From tim_one at email.msn.com  Fri Nov 26 22:48:40 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Fri, 26 Nov 1999 22:48:40 -0500
Subject: Method objects question
In-Reply-To: <slrn83lvr7.e7.bmcd@ripley.sulaco.net>
Message-ID: <000801bf388a$4b2c4ce0$742d153f@tim>

[Ben Caradoc-Davies]
> ...
> But the point is, why should functions found in a class namespace be
> converted into methods, but functions found in an instance namespace
> not converted?

Both behaviors are useful, depending on the application.  Guido picked the
behaviors for each mechanism that he judged would be most useful most often.
An obvious way to give some particular set of objects different behavior in
response to a msg is via subclassing (i.e., the class mechanism need not be
"abused" to achieve that end).

> ...
> In a nutshell, I don't see any fundamental reason why the search-
> instance-namespace then search-class-namespace attribute resolution
> order couldn't be used to allow method overriding on an
> instance-by-instance basis.

There is no "fundamental" reason.  I like it the way it is, but it's too
late to change now regardless.

> With the current, function-to-method conversion rules, this is
> impossible.

As has been explained by others, it's not impossible, just clumsy.  If the
language had been defined your way, binding an instance attribute to a plain
function would have been (at best) clumsy instead, and we'd be reading msgs
now complaining about that instead <0.4 wink>.

"the-class-determines-the-meaning-of-a-msg"-is-conventional-ly y'rs  - tim





From S.I.Reynolds at cs.bham.ac.uk  Mon Nov 29 13:07:39 1999
From: S.I.Reynolds at cs.bham.ac.uk (Stuart Reynolds)
Date: Mon, 29 Nov 1999 18:07:39 +0000
Subject: __getslice__
Message-ID: <3842C0EB.6357@cs.bham.ac.uk>

Hmm. Ok. How do slices work then?

I've written a class to do some simple structuring of data so that its
easy to plot (with, say, Gnuplot.py). The class c'tor takes:

- a sequence of input values [i1, i2, ... , in]
- a function,  f(i) --> outputi

and the objects represent the sequence:

- ( (i1, f(i1)),
    (i2, f(i2)),
     ...
    (in, f(in)),  )

Things like __repr__, __len__, __getitem__ all work as expected, but
__getslice__ behaves very oddly - which is strange since it all seems
fairly straightforward.

Any help muchly appreciated,

Stu
----------------


>>> def square(x):
        return x*x

>>> xvals = [0,1,2,3,4,5,6,7]
>>> a = UnaryFnSequence(square, xvals)
>>> xvals[1:]
[1, 2, 3, 4, 5, 6, 7]
>>> a
( (0,0), (1,1), (2,4), (3,9), (4,16), (5,25), (6,36), (7,49) )
>>> a[1:]
( (0,0), (1,1), (2,4), (3,9), (4,16), (5,25), (6,36) )

Note that the first item is still there. The last one's gone instead.

The class:


class UnaryFnSequence:
    """
    The constructor takes two arguments:
    
    fn     -- A unary function
    inputs -- An ordered sequence of inputs to the sequences,
              (i1, i2, ..., in)

    The class represents the following immutable sequence:

    ( (i1, fn(i1)),
      (i2, fn(i2)),
      ...
      (in, fn(in)),      
    )

    Slices aren't supported yet.
    """
    def __init__(self, fn, inputs):
        self._inputs = inputs
        self._fn = fn
        
    def __len__(self):
        return len(self._inputs)
            
    def __getitem__(self, index):
        x = self._inputs[index]
        return x, self._fn( x )

    def __repr__(self):
        s='( '
        for x in range( len(self) ):
            s = s+'('+`x`+','+`self._fn(x)`+'), '
        return s[:-2] + ' )'


    def __getslice__(self, low, high):
	return UnaryFnSequence(self._fn, self._inputs[low:high] )



From andy at robanal.demon.co.uk  Thu Nov  4 16:23:07 1999
From: andy at robanal.demon.co.uk (Andy Robinson)
Date: Thu, 04 Nov 1999 21:23:07 GMT
Subject: Python Conference Hotel Change
References: <199911032224.RAA10668@eric.cnri.reston.va.us>
Message-ID: <3824f510.12238077@news.demon.co.uk>

Guido van Rossum <guido at cnri.reston.va.us> wrote:

>The Foretec conference staff has this message about the upcoming
>Python conference:
>
>Due to construction delays, the Hilton Alexandria Old Town will not be
>completed in time for the Eighth International Python Conference.
>However, we are pleased to announce that we have secured an equally
>desirable facility for the conference.  The new facility is the Key
>Bridge Marriott in Arlington, Virginia.  The Key Bridge Marriott is
>conveniently located on the Virginia side of the Key Bridge in the
>Rosslyn business district with easy access to restaurants, shops, and
>the Rosslyn metro stop.  It overlooks the Potomac River, and is a 15
>minute walk across the bridge to historic Georgetown in Washington, DC.
>
>(I should add that Georgetown is, if possible, an even better location
>than old town Alexandria for the typical Python conference evening
>activities. --Guido :-)
>
My memory may be fuzzy, but isn't it an even shorter stroll from there
to Bardo, one of the world's biggest microbreweries?

- Andy




From Richard.Jones at fulcrum.com.au  Thu Nov 11 22:49:18 1999
From: Richard.Jones at fulcrum.com.au (Richard Jones)
Date: Fri, 12 Nov 1999 14:49:18 +1100
Subject: Random text generator 
In-Reply-To: Message from Rico Albanese of 1999-Nov-12 13:17:35,
            <382B86CF.AA1C12C9@qut.edu.au> 
Message-ID: <199911120349.OAA01177@envy.fulcrum.com.au>

[Rico Albanese]
> Can anyone tell me if there is a random text generator for Python?  I
> have an application that needs to generate a random password for users.
> I have looked at the random and whrandom modules but they seem to create
> random numbers only.
> 
> Thanks in advance for any advice given

   The function you want is whrandom.choice(). See

       http://www.python.org/doc/current/lib/module-whrandom.html

Python 1.5.2 (#3, Apr 19 1999, 13:39:26)  [GCC 2.8.1] on sunos5
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import whrandom, string
>>> whrandom.choice(string.letters + string.digits)
'P'
>>> whrandom.choice(string.letters + string.digits)
'e'
>>> whrandom.choice(string.letters + string.digits)
'n'
>>> whrandom.choice(string.letters + string.digits)
'I'


          Richard





From m.faassen at vet.uu.nl  Wed Nov 24 14:56:30 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 24 Nov 1999 19:56:30 GMT
Subject: automatic "lotjes trekken"
References: <81cqnv$2ip$1@vvs.superst.iae.nl> <81e3s2$j14$2@newshost.accu.uu.nl> <383BBFB9.C9607961@compaq.com>
Message-ID: <81hfte$50e$1@newshost.accu.uu.nl>

Greg Ewing <greg.ewing at compaq.com> wrote:
> Martijn Faassen wrote:
>> 
>> Not at all. everybody here speaks Dutch anyway. (it's a conspiracy!)

> Not me!

> Uunlees-iit-meens-yoo-juust-haav-too-wrijt-aal-yoor-voowls-twijce,

Oh, so you do speak Dutch as well. :)

I-understood-that-so-it-must-be-Dutch-ly yours,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?



From rree at nospam-oeb.harvard.edu  Mon Nov  1 21:31:51 1999
From: rree at nospam-oeb.harvard.edu (Rick Ree)
Date: Tue,  2 Nov 99 02:31:51 GMT
Subject: a simple helper class for parsing command line options
Message-ID: <s1sj8n.11uoa9u@dev.huh.harvard.edu>

I threw this together to make parsing command line options and arguments
a little easier than just using getopt directly.  Feedback welcome.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
from getopt import getopt, error
from string import ljust
import sys

class OptionFactory:

    """

    A class for parsing command-line options and arguments to scripts.
    See below for example code that uses the OptionFactory.

    """
    
    def __init__(self):
        self.name = None
        self.__shortstring = 'h'
        self.__longlist = ['help']
        self.__options = [('help','h','help','','print this help message')]
        self.__args = 0
        self.__argdesc = None
        self.args = None
        self.__dict = {}

    def __getitem__(self, item):
        return self.__dict[item]

    def add_options(self, options):
        """

        'options' is a sequence of tuples of the form:
        (label, short, long, value, desc)

        label = identifier for the option, used to retrieve option value
                after parsing the command line
        short = short form of option, e.g. 'h'
        long  = long form of option, e.g. 'help'
        value = descriptive string of the value the option requires
        desc  = description of the option for the help message

        """
        for option in options: self.__add(option)

    def add_argdesc(self, argdesc):
        """ add a description of the expected arguments """
        self.__argdesc = argdesc

    def __add(self, option):
        self.__options.append(option)
        label, short, long, value, desc = option
        if short:
            self.__shortstring = self.__shortstring+short
            if value: self.__shortstring = self.__shortstring+':'
        if long:
            if value:
                self.__longlist.append(long+'=')
            else:
                self.__longlist.append(long)
        self.__dict[label] = 0

    def usage(self):
        if self.__options: optstr = ' [OPTION...]'
        else: optstr = ''
        usage = 'Usage: %s%s %s\n' % (self.name, optstr, self.__argdesc)
        if self.__options: usage = '%sOptions:\n' % usage
        for option in self.__options:
            label, short, long, value, desc = option
            if short: short = '-%s' % short
            if long: long = '--%s' % long
            if value:
                if short: sval = value
                else: sval = ''
                if long:
                    lval = value; vl = '='
                else:
                    lval = ''; vl = ''
            else:
                sval = ''; lval = ''; vs = ''; vl = ''
            if short and long: comma = ',  '
            else: comma = ''
            s = '  %s%s%s%s%s%s' % (short,sval,comma,long,vl,lval)
            s = ljust(s, 30)
            s = '%s%s\n' % (s, desc)
            usage = '%s%s' % (usage, s)
        return usage
        
    def parse(self, argv):
	""" pass in sys.argv, not sys.arv[1:] """

        self.name = argv[0]
        try:
            optlist, args = getopt(argv[1:], self.__shortstring, self.__longlist)
            for opt in optlist:
                self.__process_opt(opt)
            self.args = args
        except error, x:
            msg = 'Error parsing options:  %s' % x
            sys.stderr.write('%s\n' % msg)
            sys.stderr.write('\n%s\n' % self.usage())
            sys.exit(1)

    def __process_opt(self, opt):
        option, optval = opt
        if option in ('-h', '--help'):
            print self.usage()
            sys.exit(0)
        for item in self.__options:
            label, short, long, value, desc = item
            match = 0
            if option == '-%s' % short: match = 1
            elif option == '--%s' % long: match = 1

            if match:
                if optval and value:
                    entry = self.__dict.get(label)
                    if entry:
                        entry = [entry]
                        entry.append(optval)
                        self.__dict[label] = entry
                    else:
                        self.__dict[label] = optval
                else:
                    self.__dict[label] = 1

# Run this script with options to see the OptionFactory in action

if __name__ == '__main__':

    test_flag = 'test_flag'
    verbose_flag = 'verbose_flag'
    numreps = 'numreps'
    infilename = 'infilename'
    mystery_flag = 'mystery_flag'

    opts = [(test_flag,    't', 'test',      '',     'test option'),
            (verbose_flag, 'v', 'verbose',   '',     'verbose output'),
            (numreps,      'n', '',          '#',    'do it # times'),
            (infilename,   '',  'load-file', 'FILE', 'load FILE'),
            (mystery_flag, 'x', '',          '',     'mystery option with no long version')]

    of = OptionFactory()
    of.add_options(opts)
    of.add_argdesc('[file1 file2 ...]')
    of.parse(sys.argv)

    print 'test_flag:',    of[test_flag]
    print 'verbose_flag:', of[verbose_flag]
    print 'numreps:',      of[numreps]
    print 'infilename:',   of[infilename]
    print 'mystery_flag:', of[mystery_flag]
    print 'args:',         of.args




From Andreas.Rozek at RUS.Uni-Stuttgart.De  Mon Nov 29 12:30:14 1999
From: Andreas.Rozek at RUS.Uni-Stuttgart.De (Andreas Rozek)
Date: Mon, 29 Nov 1999 18:30:14 +0100
Subject: Problem with asynch. sockets under Win98,
Message-ID: <3842B826.C91D6622@RUS.Uni-Stuttgart.De>

Hello!

I'm currently developing a distributed system which uses asynchronous
socket IO under Win98. The sockets used are set to non-blocking mode,
identified with "select" as being ready for IO and then read or writ-
ten using "recv" or "send", resp.

On the server side,  "sending"  to a TCP socket  (which has been suc-
cessfully "accept"ed before) works fine for the first time.  However,
whenever I try to "send"  to the same socket again,  I get  error no.
10045 (WSAEOPNOTSUPP, Operation not supported)

Does anybody have some experience with non-blocking socket IO for TCP
and UDP unicast and multicast? Is there an explanation for the above-
mentioned behaviour?

Basically, I'm using the following statement for writing onto the soc
ket:

  BytesSent = self.Socket.send(self.Buffer, len(self.Buffer))

(Buffer contains two characters, i.e., it's non-empty)

Thanks in advance for your effort!

-- 

Andreas Rozek
Rechenzentrum Universitaet Stuttgart       ******   **   **   *****
Kommunikationssysteme/BelWue-Entwicklung   **   **  **   **  **
Allmandring 3a                             **   **  **   **  **
D-70550 Stuttgart                          ******   **   **   *****
                                           ***      **   **       **
Telefon: ++49 (711) 685-4514               ** **    **   **       **
Telefax: ++49 (711) 678-8363               **   **   *****   ****** 
E-Mail:  Andreas.Rozek at RUS.Uni-Stuttgart.De



From jesse at student.usyd.edu.au  Wed Nov 10 08:07:29 1999
From: jesse at student.usyd.edu.au (Jesse Sweeney)
Date: Wed, 10 Nov 1999 13:07:29 GMT
Subject: IDLE and rlcompleter [Was: Re: making the Python case to my dept.]
References: <9911051740130D.02530@william.jinsa.org> <3823FC43.CF7FBD71@python.net> <m3puxno6g0.fsf@atrus.jesus.cam.ac.uk> <382523fc.14124721@news.usyd.edu.au> <m366zcoagc.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <382960a5.14921242@news.usyd.edu.au>

On 08 Nov 1999 23:00:35 +0000, Michael Hudson <mwh21 at cam.ac.uk> wrote:

>jesse at student.usyd.edu.au (Jesse Sweeney) writes:
>
>> Coincidentally, I had the same idea just a few days ago, and after a
>> bit of tinkering, I came up with two different IDLE extensions. I've
>> put the files on the web at
>> 
>> http://www-personal.usyd.edu.au/~jsweeney/python .
>
>Funky!
>
>I had to change <Meta-.> to <Meta-period> and then to <Alt-period>
>(the latter is probably a misconiguration at this end, but the first
>may not be).

Thanks, Michael, I changed this, and fixed a few other things that I thought
needed fixing, and uploaded the files again.

>I'm very used to some other readline conventions such as Meta-. being
>yank-last-arg and C-r being reverse-search-history and some
>others. Would these be easy to add to your extension? (I know next to
>nothing about Tkinter).

I doubt I know much more, but that's no excuse! I'm going on the IDLE source
code, extend.txt, the several extension examples that come with IDLE and
Fredrik's Introduction to Tkinter whenever I need it (I had to look up the staff
about text marks, for instance).

Now, yank-last-arg requires knowing about the history, and extensions --  as far
as I can figure -- don't get told about the history. So, that means this
requires editing the IDLE source, rather than making an extension. Here's what I
came up with (sorry, I don't have a working copy of patch here): 

1. Add a yank_last_arg method to the History class in IdleHistory.py:

    def yank_last_arg(self, event):
        if len(self.history) == 0:
            self.text.bell()
            return "break"
        item = self.history[-1]
        wordchars = string.letters + string.digits + "_" + "."
        i = len(item)
        while i > 0 and item[i-1] in wordchars:
            i = i-1
        last_arg = item[i:]
        self._put_source("insert", last_arg)
        return "break"


2. Bind the event in History.__init__ by adding a line of the form:

        text.bind("<<yank-last-arg>>", self.yank_last_arg)


3. Bind a keystroke of your choosing to the function by adding a line to
keydefs.py like:

	'<<yank-last-arg>>': ['<Alt-period>'],

(4. Edit Bindings.py (I think) if you want a menu item as well.)

That should do it. I'll put my copy of IdleHistory.py on my website (the address
is above) to make it more straightforward.

reverse-search-history shouldn't be *that* much harder to implement, except that
I haven't actually worked out how to capture the shell input yet. Nevertheless,
notice that I'm not putting my hand up for the job. I've just noticed that the
IDLE todo file (TODO.txt) mentions this:

- more emacsisms:
    ...
    - incremental search?

So perhaps Guido will save us all the trouble, and leave me to study for my
exams ;).

>> Obviously, unlike "Expand Word" these only work in the Python shell,
>> not in the file-editing windows. (Now that I think about it, this
>> would probably be pretty useful in those editing windows, but that
>> would take some partial compilation of some sort I suppose, which is
>> way too much work for me.)
>
>I think I'd settle for the completions being selected from the words
>known to the Python shell, if that helps.

Well, by a happy coincidence this is how it works at the moment. So
AutoComplete.py works quite nicely in that regard, but RLComplete.py is pretty
useless, because it just prints the possible completions to stdout, which is the
Python shell if you've got one open, and I suppose you could kinda put the
windows next to each other, but 'urk' is all I can say to that. A CallTips-style
thing could be hacked together I suppose, if someone was dying for this
functionality, but I doubt it's worth the effort.

>Good work!
>
>Cheers,
>Michael

Thanks!

	Cheers, Jesse.



From pnagel at tbs.co.za  Thu Nov 18 05:30:23 1999
From: pnagel at tbs.co.za (Pieter Nagel)
Date: Thu, 18 Nov 1999 12:30:23 +0200 (SAST)
Subject: Beginning programming with Python
In-Reply-To: <383195C1.826B3342@iqsoft.hu>
References: <sng0ychgwt.fsf@motorola.com> <382B3A53.5816B2@callware.com> <382B3D79.74CD3E0F@email.sps.mot.com> <80p9ds$ppo$1@newshost.accu.uu.nl> <asteroidal-80psfo/INN-2.2.1/bat@broadway.news.is-europe.net> <38312D55.A36EB56D@iqsoft.hu> <antigorite-80rol2/INN-2.2.1/attentive@broadway.news.is-europe.net> <383195C1.826B3342@iqsoft.hu>
Message-ID: <Pine.LNX.4.10.9911181226130.1914-100000@iguana.home.tbs.co.za>

On Tue, 16 Nov 1999, Nemeth Miklos wrote:

> I am neither a physicist nor a mathematician, I am a simple business software
> engineer. I like Python but I'd like to be convinced that Python is not only
> for scientific programming, but also for business programming.

Strange, I see things the other way around.

I think Python is great for business software, but I would think that
the scientific community would need some convincing that Python is
good for scientific software too.

Of course, to me, scientific software conjures up images of
mathematic-intense software. Stuff like Fortran and Matlab. 

Or intensely paralel weather-prediction software written in C running
on Beowulf clusters.

Hmm. Even in *that* scenario python would be great for tying the
whole together, even if the C code is responsible for the actual
megaflops.

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

    Tangible Business Software
    http://www.tbs.co.za
    +27 12 46 8563	




From ivnowa at hvision.nl  Mon Nov  8 16:40:11 1999
From: ivnowa at hvision.nl (Hans Nowak)
Date: Mon, 8 Nov 1999 22:40:11 +0100
Subject: Popup menus in Tkinter
Message-ID: <199911082139.WAA02023@axil.hvision.nl>

On 8 Nov 99, dkaznadzey at my-deja.com wrote:

> Thanks A LOT !
> It worked perfectly!
> What would be a good source of information on tk commands
> available through Tkinter?

Myself, I'm just using the Tkinter introduction by Fredrik Lundh. It
doesn't cover everything (that's why it's an introduction I suppose  
;-) but usually it's sufficient. For me, that is. A book would be 
nice though.  ^_^

--Hans Nowak (zephyrfalcon at hvision.nl)
Homepage: http://fly.to/zephyrfalcon
You call me a masterless man. You are wrong. I am my own master.



From fdrake at acm.org  Tue Nov 16 14:40:22 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Tue, 16 Nov 1999 14:40:22 -0500 (EST)
Subject: Final classes like in java???
In-Reply-To: <80sao7$1848$1@rtpnews.raleigh.ibm.com>
References: <80sao7$1848$1@rtpnews.raleigh.ibm.com>
Message-ID: <14385.45862.55010.146256@weyr.cnri.reston.va.us>

Arinte writes:
 > Is there a way you can define your python class to be final or not able to
 > be inherited from?

  Not guaranteed, but this should be good enough:

------------------------------------------------------------------------
class FinalClass:
    def __init__(self):
        if self.__class__ is not FinalClass:
            raise SomeError, "you can't subclass from me!"
------------------------------------------------------------------------

  Whether or not this is a good idea is another thing.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From quinn at amatho.ugcs.caltech.edu  Tue Nov 30 17:25:42 1999
From: quinn at amatho.ugcs.caltech.edu (Quinn Dunkan)
Date: 30 Nov 1999 22:25:42 GMT
Subject: performance of tuple-less Python?
References: <1268226961-30285235@hypernet.com> <943917954snz@vision25.demon.co.uk> <slrn8471sd.m5h.nascheme@adler.dynodns.net>
Message-ID: <slrn848jos.nnk.quinn@amatho.ugcs.caltech.edu>

On Tue, 30 Nov 1999 08:14:36 GMT, Neil Schemenauer <nascheme at enme.ucalgary.ca>
wrote:
>Phil Hunt <philh at vision25.demon.co.uk> wrote:
>>In article <1268226961-30285235 at hypernet.com>
>>           gmcm at hypernet.com "Gordon McMillan" writes:
>>> They're only redundant if the concept of mutability is 
>>> meaningless. A reference-based language without a concept 
>>> of mutability would be, um, an adventure.
>>
>>Why?
>
>Some values should be immutable.  Consider if integers in Python
>were mutable.  You store a reference to 1 and somewhere else the
>value of 1 is changed to 2.  Is that not a problem?  Also, you
>can not (efficiently) implement hash tables.
>
>Maybe you mean that mutability is not required.  That would be an
>interesting language.  How to you accomplish anything?  Sometimes
>you want mutability and sometimes you don't.  That is why both
>tuples and lists exist.

See any purely functional language, like haskell.  You can simulate mutability
in haskell by passing monads around, but you might be surprised at how much
you can get done with no side effects.

And what was that party trick of changing 1 to equal 2 in forth?  Maybe that's
more at the language level though, but I don't know much about forth (one of
these days...).



From a.eyre at optichrome.com  Mon Nov 29 12:30:05 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Mon, 29 Nov 1999 17:30:05 -0000
Subject: Python bugs
In-Reply-To: <51C9F3C392D0D111A15600A0C99E64A9AF5690@localhost>
Message-ID: <001101bf3a8f$5fe300f0$3acbd9c2@peridot.optichrome.com>

> >>> import shelve
> >>> s = shelve.open('foo')
> >>> s.get('bar', 'baz')
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
> AttributeError: get

This one annoyed me more. Probably not fixable, but...

>>> s = shelve.open("foo")
>>> l = []
>>> s["bar"] = l
>>> s["bar"]
[]
>>> l.append("baz")
>>> s["bar"]
[]

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From gerrit.holl at pobox.com  Fri Nov 19 07:45:43 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Fri, 19 Nov 1999 13:45:43 +0100
Subject: Beginning programming with Python
In-Reply-To: <80s22o$o0l$1@nntp9.atl.mindspring.net>
References: <sng0ychgwt.fsf@motorola.com> <80p9ds$ppo$1@newshost.accu.uu.nl> <asteroidal-80psfo/INN-2.2.1/bat@broadway.news.is-europe.net> <38317787.19AFDAF5@compaq.com> <80s22o$o0l$1@nntp9.atl.mindspring.net>
Message-ID: <19991119134543.A4278@optiplex.palga.uucp>

Aahz Maruch wrote:
> >Well, not quite - you can forget the "self" in method
> >declarations fairly easily...
> 
> I don't seem to have that problem; my fingers have learned to
> automatically type "(self" whenever "def" is indented.  ;-)

Hmm, don't you ever define a function within an IF?

if __name__ == '__main__':
	def main():
		return 'main'
	main()

But that's ugly indeed...

And what about defs in defs?

regards,
Gerrit.

-- 
"We all know Linux is great...it does infinite loops in 5 seconds."
(Linus Torvalds about the superiority of Linux on the Amterdam
Linux Symposium)



From thor at localhost.localdomain  Sun Nov 14 19:18:57 1999
From: thor at localhost.localdomain (Manuel Gutierrez Algaba)
Date: 15 Nov 1999 00:18:57 GMT
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <JlFX3.3659$s87.67930@dfiatx1-snr1.gtei.net>
Message-ID: <slrn82unnr.mja.thor@localhost.localdomain>

On Sun, 14 Nov 1999 20:56:09 GMT, Jules Allen
 <usenet-mm-yyyy at julesallen.com> wrote:

>Thanks for the History lesson! This helps put it all in perspective
>but why *something* hasn't been built before now is kind of weird.

The problem we have here, it's that we need a easy method of putting
many information together, so it's easy for the _newbie_ and for
the wizard to find it.
Trove has proved to be a failure just because it's too complex
to programm it, to finish it. Why we don't attach to the simplest
method of collecting info ?

Why don't we make small pieces of info attributed like this:
\jiji
BabylMailbox

Access a Babyl mailbox, which is similar to an
MMDF mailbox.  Mail messages start with a
 line containing only <literal>'*** EOOH ***'</literal> and end
 with a line containing only <literal>'\037\014'</literal>.

 A file object fp that  points to the mailbox file.

Initialize the mailbox object and point to the first
 message in the mailbox.

\indexmoduleinbox  \indexBabylMailbox \indexclassdefinition
\indexmmdf
\jiji    

This is so general and vague that can be used for ANYTHING,
and it's so simple that anybody can use it. This is 
the method i've used for:
http://www.ctv.es/USERS/irmina/TeEncontreX.html

Why to make it complex if it can be very simple ???
Only simple and general things are commonly accepted and used
(or it should be so!)
-- 
Manolo



From bernhard at alpha1.csd.uwm.edu  Mon Nov  8 10:44:02 1999
From: bernhard at alpha1.csd.uwm.edu (Bernhard Reiter)
Date: 8 Nov 1999 15:44:02 GMT
Subject: Graphics.
References: <3826E949.BBA1BAFD@imf.au.dk> <3826ED30.70FBE3D3@callware.com>
Message-ID: <slrn82dru1.p5o.bernhard@alpha1.csd.uwm.edu>

On Mon, 08 Nov 1999 08:33:04 -0700, Ivan Van Laningham <ivanlan at callware.com> wrote:

>> How do I draw some graphics on a screen using Python running on
>> HP-UX
>> MacOS
>> Windows

>Use the Tkinter Canvas widget.  Start at
>	http://www.pythonware.com

Other choices include the python bindings for
	wxWindows (wxPython)
and	Fox 	(pyFox?)

	Bernhard


-- 
Research Assistant, Geog Dept UM-Milwaukee, USA.    (www.uwm.edu/~bernhard)
Free Software Projects and Consulting 			   (intevation.net)  
Association for a Free Informational Infrastructure              (ffii.org)



From webmaster at python.org  Tue Nov 16 14:25:52 1999
From: webmaster at python.org (Python.Org Webmaster)
Date: Tue, 16 Nov 1999 14:25:52 -0500 (EST)
Subject: Planned power outage tonight at CNRI
Message-ID: <14385.44992.984861.882508@anthem.cnri.reston.va.us>

Virginia Power has informed us that the power to CNRI's building will
be out between 2:00am and 6:00am Wednesday November 17th, EST.  This
will affect all Python.Org machines and services.  Apologies for any
inconveniences.

-Barry



From shouldbe at message.com  Mon Nov 29 17:01:54 1999
From: shouldbe at message.com (Arinte)
Date: Mon, 29 Nov 1999 17:01:54 -0500
Subject: Python Path weirdness???
Message-ID: <81usqn$1lou$1@rtpnews.raleigh.ibm.com>

I have this code in my embedded app

 char* path ="D:\\POSSApps\\anotherposs app\\Embed Python\\some.exe";
 Py_SetProgramName(path);
 Py_Initialize();
 fname = fopen(name,"r");
 Py_InitModule("poss", Exposem);
 cout<<Py_GetPath()<<endl;

Some.exe is not the name of my program, but I want python to look in
different directory for .pyc files.
Py_GetPath() returns a string that does not include
"D:\\POSSApps\\anotherposs app\\Embed Python\\some.exe".

Is there another way to get this to work through my c program?  Or should
this be working?

Thanks.





From usenet-mm-yyyy at julesallen.com  Tue Nov  9 16:00:47 1999
From: usenet-mm-yyyy at julesallen.com (Jules Allen)
Date: Tue, 09 Nov 1999 21:00:47 GMT
Subject: [Python] header (fwd)
References: <209f01bf2ac7$c19b1d50$f0c809c0@lslp7o.lsl.co.uk>
Message-ID: <3Y%V3.438$QY4.16745@dfiatx1-snr1.gtei.net>

| ...problem to those (for instance) who are unlucky enough to have to use
| Outlook Express...

Why not subscribe via an alias at iname.com and just use it for the
Python list. It'll give you a solid hook to filter on. Or get a
deja.com, etc., freebie email address and check there?

Good luck!

-- 
"People," said Oscar Wilde, "are either perfectly charming or
 perfectly tedious."




From aa8vb at yahoo.com  Tue Nov  2 13:22:07 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Tue, 2 Nov 1999 13:22:07 -0500
Subject: Tkinter canvas blow-up:  bbox(ALL) == None
In-Reply-To: <19991102072202.A134170008@vislab.epa.gov>
References: <19991101102103.A134035786@vislab.epa.gov> <013c01bf2510$57457f20$f29b12c2@secret.pythonware.com> <19991102072202.A134170008@vislab.epa.gov>
Message-ID: <19991102132207.A134181398@vislab.epa.gov>

Randall Hopper:
 |Fredrik Lundh:
 | |Randall Hopper <aa8vb at yahoo.com> wrote:
 | |>      Apparently if the bounds of the items in a Tk canvas widget exceed
 | |> MAXINT, bbox(ALL) returns None.  The attached Python script demonstrates
 | |> this.
...
 |Ok, thanks for the info.  So do you know if Tk uses ints internally as
 |well, or is its returning the empty string an artifact of how it was fed by
 |Tkinter?

Sorry.  Too many pronouns and implied nouns in that sentence.  Should read:

 So do you know if Tk uses ints internally [to represent the bounding box],
 or is Tk's returning the empty string an artifact of how Tk was fed by
 Tkinter?

Randall



From tbryan at python.net  Tue Nov 30 19:16:45 1999
From: tbryan at python.net (Thomas A. Bryan)
Date: Tue, 30 Nov 1999 19:16:45 -0500
Subject: Python complaints
References: <000701bf36c0$cdb9f480$292d153f@tim>
Message-ID: <384468ED.7CA86357@python.net>

Tim Peters wrote:
> 
> [Thomas A. Bryan]
> > ...
> > For example, I saw some code that I wanted to copy in DejaNews
> > c.l.py archives, but the lines wrap in narrow columns in DejaNews.
> > I had to reindent the code.  Grrr.
> 
> In DejaNews, while looking at a msg, click the "View original Usenet format"

Thanks to everyone who pointed that out.

> > I have also had problems when e-mailing code when I e-mail the
> > code as text in the message instead of attaching a document.
> >
> > Is there a better way to transmit code so that possible
> > indentation munging won't force the reader to reindent?
> 
> If you don't want to use attachments (chicken <wink>), 

I normally do....but when I'm sending e-mail from a remote machine, 
I sometimes get lazy and cut and paste from my local machine. :-)

---Tom



From gmcm at hypernet.com  Tue Nov 16 16:00:23 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 16 Nov 1999 16:00:23 -0500
Subject: Final classes like in java???
In-Reply-To: <80sao7$1848$1@rtpnews.raleigh.ibm.com>
Message-ID: <1269336468-10034142@hypernet.com>

Arinte writes:

> Is there a way you can define your python class to be final or not able to
> be inherited from?

You could deter subclassing, but not prevent it (I can always 
rebuild my class's __bases__ to include your Final class).

Final in Java is there for efficiency. There would be no gains in 
Python since the compiler has no role in choosing the override.

- Gordon



From rkaiser at wwa.com  Sun Nov 28 17:18:32 1999
From: rkaiser at wwa.com (Ralph Kaiser)
Date: Sun, 28 Nov 1999 16:18:32 -0600
Subject: IDLE & Linux
References: <B466ECEC.F8E%Dan@Grassi.com>
Message-ID: <uQh04.4688$tI2.97781@ord-read.news.verio.net>

Dan Grassi <Dan at Grassi.com> wrote in message
news:B466ECEC.F8E%Dan at Grassi.com...
> I am trying to get IDLE to run.  The docs say that it is automatically
> installed but that is just not true.  There is a /usr/bin/idle but when it
> runs it quits complaining that it can't find
> /usr/lib/python1.5/site-packages/idle/idle.py

Hi,

/usr/bin/idle is just a one line script to run idle.py Use "find .-name
idle.py"
to locate it. on my system it's
/usr/local/src/Python-1.5.2/Tools/idle/idle.py.
Good Luck

Ralph





From mwh21 at cam.ac.uk  Fri Nov 26 15:59:02 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 26 Nov 1999 20:59:02 +0000
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <867lj85g4c.fsf@g.local> <al8yabo9fm3.fsf@sirppi.helsinki.fi> <86so1v4jlj.fsf@g.local> <al8hfiaan8p.fsf@sirppi.helsinki.fi> <86ln7m2t31.fsf@g.local>
Message-ID: <m3emddrmuh.fsf@atrus.jesus.cam.ac.uk>

Gareth McCaughan <Gareth.McCaughan at pobox.com> writes:

> Markus Stenberg wrote:
> 
> > I have always found lisp-ish syntax of "masses of ()s" disconcerting in
> > large amounts - admittedly, I've written more than few lines of
> > [ce]lisp/scheme/.., but never enjoyed browsing the results very much. I
> > consider Dylan to be "readable" clisp.
> 
> One quickly gets used to Lisp syntax. Like using indentation to
> denote block structure. :-) After a while, you just stop noticing
> all those parentheses.

This leads to a thought: would it be possible using lisp's silly
amount of flexibility to get the reader to auto-paranthesize (word?
doubt it):

defun our-member (obj lst)
  if null lst
     nil
     if eql (car lst) obj
        lst
        our-member obj (car lst)

based on indentation, as

(defun our-member (obj lst)
  (if (null lst)
    nil
    (if (eql (car lst) obj)
      lst
      (our-member obj (car lst)))))

or is whitespace ignorance built-in?

indentation-as-structure-*will*-rule-the-world-ly y'rs
Michael




From ayinger1 at my-deja.com  Tue Nov  9 22:51:48 1999
From: ayinger1 at my-deja.com (yinger)
Date: Wed, 10 Nov 1999 03:51:48 GMT
Subject: Detect file existance before open?
References: <zWaO3.2925$lv2.105561@news.uswest.net> <380c92a7.412580789@192.77.87.74> <80a29u$gn9$1@nnrp1.deja.com>
Message-ID: <80aq4j$28q$1@nnrp1.deja.com>

import os.path
os.path.isfile('<filename>')

will also work.

> import commands
> commands.getstatus('<filename>')
>
> In article <380c92a7.412580789 at 192.77.87.74>,
>   jlj at cfdrc.com (Lyle Johnson) wrote:
> > os.path.exists(filename)
> >
> > On Sat, 16 Oct 1999 21:06:07 -0600, "Bill Rodgers" <Bill at nospam.com>
> > wrote:
> >
> > >I would like to check for the existance of a file before I attempt
an
> open
> > >and maybe throw an exception.  I don't see a way to do this.  I
would
> like
> > >to tell the user that the file he specified doesn't exist.  Is
there
> a way
> > >to do this or do I need to catch the exception and deal with it
that
> way?
> > >Thanks,
> > >--
> > >Bill Rodgers
> > >Enterprise Management Consultant
> > >Specialty: BMC/Patrol
> > >brodg1 at ix.netcom.com
> > >
> > >
> >
> >
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
>


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



From webmaster at python.org  Mon Nov 15 12:18:23 1999
From: webmaster at python.org (webmaster at python.org)
Date: Mon, 15 Nov 1999 12:18:23 -0500 (EST)
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net>
	<3d1z9sq2uc.fsf@amarok.cnri.reston.va.us>
Message-ID: <14384.16479.836502.161209@anthem.cnri.reston.va.us>

>>>>> "AMK" == Andrew M Kuchling <akuchlin at mems-exchange.org> writes:

    AMK> http://www.vex.net/~x/parnassus/ is *very* nice, and it might
    AMK> be the only solution required.  Apparently a good deal of
    AMK> work has been put into seeding the initial database, and
    AMK> that's important, because it makes the index immediately
    AMK> useful.  Yet I hadn't heard about it until now; is the site
    AMK> still beta?  Did I miss a c.l.py.announce article about it?

    AMK> There are various issues with it -- I don't know how scalable
    AMK> the underlying code is, the graphic design could be simpler,
    AMK> vex.net's reliability and bandwidth are unknown, and we can
    AMK> argue about the classification hierarchy *forever* -- but
    AMK> it's still valuable.

    AMK> We need to just pick a solution and stick with it; Parnassus
    AMK> is as good as any, and has the incalculable advatange of
    AMK> being implemented.  I'd like to encourage everyone who
    AMK> maintains Python code, or writes Python tutorials and docs,
    AMK> to go to http://www.vex.net/~x/parnassus/ .  Is everything
    AMK> you've done in the index?  If not, add the missing items.
    AMK> Can we prominently link to it from python.org, so people
    AMK> begin to use it?

Very interesting!  I hadn't heard about it until Andrew's post either.
I just made an update to the Mailman entry, but it looks like there's
some editorial review for record changes, because the change isn't
effective immediately.  That could be another bottleneck in the
system.

In any event, since the site's been outed, I'll probably add a link
somewhere on the python.org site, but I'd still like Parnassus' author
to get in touch with me...

-Barry



From donn at u.washington.edu  Tue Nov 23 17:40:25 1999
From: donn at u.washington.edu (Donn Cave)
Date: 23 Nov 1999 22:40:25 GMT
Subject: dir() and propertly list NEWBIE
References: <81eorl$b60$1@nnrp1.deja.com>
Message-ID: <81f54p$ig4$1@nntp6.u.washington.edu>

Quoth tiddlerdeja at my-deja.com:
| Is it possible to get the functions of an object using dir()? Rather
| that looking them up in a book/API.

Yes, though not directly through the object and it's not real simple.

   def printcdir(c):
       print dir(c)
       for b in c.__bases__:
           printcdir(b)

   def printxdir(instance):
       print dir(instance)
       printcdir(instance.__class__)

| Similarly, If I can get an objects functions, is it possible then get a
| list of the possible parameters for that function?

No!  Wouldn't that be nice?

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu



From robin at jessikat.demon.co.uk  Thu Nov 18 19:46:31 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Fri, 19 Nov 1999 00:46:31 +0000
Subject: Comparing Python to (free) Math languages/environments
References: <19991118014716.13748.00000334@ng-bj1.aol.com>
 <Pine.LNX.4.10.9911180305320.7208-100000@us2.mayo.edu>
Message-ID: <0MjAOJAn3JN4EwHI@jessikat.demon.co.uk>

In article <Pine.LNX.4.10.9911180305320.7208-100000 at us2.mayo.edu>,
Travis Oliphant <olipt at mayo.edu> writes
...
>
>I've been steadily working over the past year to flesh out some of the
>missing numerical pieces (see http://oliphant.netpedia.net).  As to
>how things are coming, I can only give my own experience.  With the
>packages I've put together I am able to use Python exclusively in my day
>to day work and find that I am able to do things I would never have tried
>in MATLAB.  
>
>In short, I do not regret the decision to use Python.  In fact I'm happier
>about the decision each day.
>

me too

> on the lists.  Good luck in your search.
>
>Travis Oliphant
>

-- 
Robin Becker



From doughellmann at home.com  Mon Nov  1 08:33:55 1999
From: doughellmann at home.com (Doug Hellmann)
Date: Mon, 01 Nov 1999 13:33:55 GMT
Subject: Tkinter app 'freezes' when I use after_idle
References: <ksu2na636b.fsf@telelogic.se> <3819864B.297FCB9E@home.com> <ks7lk653oo.fsf@telelogic.se>
Message-ID: <381D9811.83654620@home.com>


Andrew Markebo wrote:
> 
> | network, it might take a while to come back.  Keep in mind that as long
> | as the program is running your callback code, the display will not be
> | updated.
> 
> *HMM* Ok lets try to sort this out.. Of course the first call to the
>  function is from a callback, when I press the 'go' button a call is
>  made to 'checkfunc'. Last in checkfunc I call
>  self.after_idle(checkfunc) this would stay away from the event??


An overly simplified look at the process flow for a Tk app might look like:

create widgets
enter mainloop:
    look for events (including update)
    if events: 
        call event callbacks (your code)
    else: 
        call idle handlers (your code)


While you are in a callback or idle handler, program control does not go
through the section that allows the mainloop to dispatch a callback. 
Instead, the events that would cause the callback to be called are being
queued up for the next time around the mainloop.

It is possible to shortcut this flow somewhat by using the
update_idletasks() method to make sure that all of the screen painting
calls are made (to handle exposure events).

What you want to try to do with your idle handler is make it run very
quickly, so that you can get in and out before a lot of events start
queuing up and the user notices that the screen is not refreshing. 
After you are done with your processing, you reregister the idle handler
to run again.

> |
> | You could use a file handler to let Tk watch the socket for you (instead
> | of doing your own select).  You would then get a callback when there is
> | data for you to read from the socket.  I've had success with this scheme
> 
> *bonk* I have heard about that, time to take a dig into my TCL/TK book again..

That may be your best bet for this app.  There are cases when doing your
own call to select might make more sense.

Doug



From da at ski.org  Tue Nov  2 14:37:53 1999
From: da at ski.org (David Ascher)
Date: 2 Nov 1999 13:37:53 -0600
Subject: [CORRECTED] Dr. Dobb's Python-URL! - weekly Python news and links (Nov  2)
Message-ID: <C6ABBD6A13831F60.4B642F6732EFD180.B5584DB172F56E61@lp.airnews.net>

Today is election day in the United States.  Here are the winners from a
much simpler yet equally non-scientific selection process:

    The Python Consortium is started!  Hewlett-Packard and LLNL provides
    the marquee status and the big bucks, Digital Creations the wildly
    successful application, Interet quietly chips in, along with Foretec
    Seminars, the first organization that had a check in hand!
        http://www.python.org/consortium/

    Michel Pelletier (of Digital Creations) and Dan York (of LinuxCare)
    announce that they are writing a Zope book for O'Reilly and Associates:
        http://www.zope.org/Members/michel/ZopeBook/index_html/index_html
  
    Jan Walter announces some scripts for Blender, a computer graphics
    rendering tool, with very nice pictures to boot:
        http://www.q-bus.de/Blender/

    David S. Harrison puts forth code which formats number for human
    consumption:
        http://www.deja.com/getdoc.xp?AN=542270123

    Scot Hacker explains how good scripting BeOS is
	http://www.byte.com/column/BYT19991026S0001

    Tim Peters sort of sorts out some of the sort issues sorted out
    between 1.5.1 and 1.5.2:
        http://www.deja.com/getdoc.xp?AN=541957314
  
    The same Tim accounts for a surprising 'feature' involving testing
    the object identity of immutable objects which puzzled Pearu
    Peterson and yours truly:
        http://www.python.org/pipermail/matrix-sig/1999-November/003106.html


=========================================================================
Everything you want is probably one or two clicks away in these pages:

      Python.org's Python Language Website is the center of Pythonia
          http://www.python.org

      Much of Python's real work takes place on Special-Interest Group
          mailing lists http://www.python.org/sigs/

      Python Consortium emerges as an independent nexus of activity
          http://www.python.org/consortium

      Python To-Do List anticipates some of Python's future direction
          http://www.python.org/cgi-bin/todo.py

      Python Journal is at work on its second issue
          http://www.pythonjournal.com

      Links2Go is a new semi-automated link collection; it's impressive
      what AI can generate
          http://www.links2go.com/search?search=python

    Archive probing trick of the trade:
        http://www.dejanews.com/dnquery.xp?QRY=&DBS=2&ST=PS&defaultOp=AND&LNG=ALL&format=threaded&showsort=date&maxhits=100&groups=comp.lang.python


Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://purl.org/thecliff/python/url.html
or
  http://www.dejanews.com/dnquery.xp?QRY=~g%20comp.lang.python%20Python-URL%21


Suggestions/corrections for next week's posting are always welcome.
  http://www.egroups.com/list/python-url-leads/

To receive a new issue of this posting in e-mail each Monday morning,
ask <claird at neosoft.com> to subscribe.  Mention "Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.
-- 

Cameron Laird           http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX



From thomas.heller at ion-tof.com  Tue Nov 30 14:51:59 1999
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Tue, 30 Nov 1999 20:51:59 +0100
Subject: Exposing COM via XML-RPC or Something Else
References: <613145F79272D211914B0020AFF6401914DD7C@gandalf.digicool.com>
Message-ID: <00f001bf3b6c$5debe270$4500a8c0@thomasnotebook>

> You might want to take a look at MS' SOAP (Simple Object Access 
> Protocol) spec - it's not fully baked yet, but it basically takes
> xml-rpc several steps further, allowing you to communicate with 
> SOAP-enabled objects (COM or otherwise) using XML over HTTP. We in 
> Zope-land are keeping a close eye on this, as publishing objects 
> via HTTP is the core idea of Zope already.

Zope and SOAP sounds very similar.
Is this by accident?

Thomas Heller




From mlh at vier.idi.ntnu.no  Tue Nov 16 15:16:48 1999
From: mlh at vier.idi.ntnu.no (Magnus L. Hetland)
Date: 16 Nov 1999 21:16:48 +0100
Subject: list.without()?
References: <11A17AA2B9EAD111BCEA00A0C9B41793034AAEE7@molach.origin.ea.com>
Message-ID: <y0jbt8ui43z.fsf@vier.idi.ntnu.no>

"Stidolph, David" <stidolph at origin.ea.com> writes:

> so ...
> 
> def without(source, element):
>   while element in source:
>     source.remove(element)  
> 
> Is that what you are looking for?

No - but thanks anyway ;)

What I wanted was a nondestructive method included in the language.
According to some (i.e. Frank Drake IIRC) that is contrived, and I
have conceded on the issue. Implementing it myself is not a problem
(*especially* now that I have received numerous implementations from
other people... ;)

> It would help if I had read the earlier
> posts!  :)

--

  Magnus          Echelon jamming noise:
  Lie             FBI CIA NSA Handgun Assault Bomb Drug Terrorism
  Hetland         Special Forces Delta Force AK47 Hillary Clinton 



From jam at quark.emich.edu  Tue Nov 30 08:03:26 1999
From: jam at quark.emich.edu (Jeff)
Date: Tue, 30 Nov 1999 08:03:26 -0500
Subject: Looking for sshlib.py
In-Reply-To: <iwI04.9621$VA6.47863@news-server.bigpond.net.au>
References: <iwI04.9621$VA6.47863@news-server.bigpond.net.au>
Message-ID: <19991130080326.A1376@quark.emich.edu>

On Tue, Nov 30, 1999 at 04:37:34AM +0000, John Leach wrote:
> Is there an ssh lib for Python. Or anything similar. I'm looking for
> something more secure than telnet.
> Any ideas? Write my own?
> John Leach
> http://www.bosware.com.au
> 

greetings,

what type of application are you trying to write that you need this type of
access? if all you want is a 'secure shell' so you can connect to a remote
machine, ssh is the way to go.. I don't know if you can hook it into python
though.

regards,
J
-- 
|| visit gfd <http://quark.emich.edu/>
|| psa member #293 <http://www.python.org/> 
|| New Image Systems & Services, Inc. <http://www.newimage.com/>



From robin at jessikat.demon.co.uk  Fri Nov  5 09:44:26 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Fri, 5 Nov 1999 14:44:26 +0000
Subject: [OFFTOPIC] Maps and functions (was Re: Apply a function to each list member?)
References: <slrn824gee.1f7.bwinton@tor.dhs.org>
 <XFMail.991105110617.mikael@isy.liu.se> <19991105215504.A497@Ridcully.home>
 <14370.53219.667492.946075@gargle.gargle.HOWL>
Message-ID: <27NwjKAK1uI4EwdJ@jessikat.demon.co.uk>

In article <14370.53219.667492.946075 at gargle.gargle.HOWL>, Charles G
Waldman <cgw at fnal.gov> writes
>Malcolm Tredinnick writes:
> > 
> > Correct. In fact, since we're being precise, let me point out that
> > (mathematically speaking), all functions are maps (or mappings), but not
> > all mappings are functions. The point is that a function should have only a
> > single "output" for any given input (i.e. it should be many-to-one or
> > one-to-one, not one-to-many). So some mapping that sends, say, the number 2
> > to the numbers 3 and 4 is not a function (but a mapping that sends both the
> > numbers 3 and 4 to the number 2 is a function, since for each input there
> > is only a single output).
>
>Hmm... what kind of "map" are you thinking of that could send 2 to
>"both 3 and 4"?  I think such a thing is not a mapping, at least not
>according to any definition of "map" that I've ever seen.  If you
>look, for instance, in any basic topology text, you will see that
>"map" and "function" mean exactly the same thing, when used as nouns.
>"Map" also has a sense as a transitive verb, which means "to apply a
>function to", which is where the Python/functional programming usage
>comes from.
>
>
>
I define f as a thing which takes integer arguments and outputs sets of
integers

presumably I can have such an f that

        f(2)={3,4}

to my simple engineer's eyes f could belong to the functions mapping Z--
>2^Z

I assume we're not talking that way if the f above isn't a function.
-- 
Robin Becker



From mhammond at skippinet.com.au  Mon Nov 15 06:51:39 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Mon, 15 Nov 1999 11:51:39 GMT
Subject: Python ISAPI
References: <80oq15$ln8$1@oceanite.cybercable.fr>
Message-ID: <ftSX3.4566$Xo.24188@news-server.bigpond.net.au>

AFAIK there is no ISAPI module - but Python is an AXScript language, so it
works fine in ASP (hence no real demand for ISAPI)

Mark.

Florent Rami?re wrote in message <80oq15$ln8$1 at oceanite.cybercable.fr>...
>Hello,
>
>    i am looking for an python ISAPI, but could not find any.
>    Could you tell me where i can get such a goodie ?
>    Does anybody ever worked on this ?
>
>Thanks.
>
>





From usenet-mm-yyyy at julesallen.com  Wed Nov 24 10:34:45 1999
From: usenet-mm-yyyy at julesallen.com (Jules Allen)
Date: Wed, 24 Nov 1999 15:34:45 GMT
Subject: Python books
References: <199911230613.HAA21179@dionysus.fw.cuci.nl> <383A9EF4.5D1E8986@callware.com> <00a901bf35c3$2549ed80$f29b12c2@secret.pythonware.com>
Message-ID: <pAT_3.271$lT.6459@dfiatx1-snr1.gtei.net>

| well, the eff-bot guide is definitely an annotated script
| collection, but I think Hans was referring to Martin Brown's
| upcoming pMatter book; [snip]

I really enjoyed Martin's Perl book and I'm looking forward to this
one. He's got an easy to read style, not unlike Daryl H/Ken McD's
Quick Python Book. Amazon claimed they shipped me a PyAnnotated
last week but I've yet to see it.

Personally I think the Quick Python book is one of the better
technical books I've ever read. The last books to have that much
impact on my so-called-career is Cooper's semi technical "Inmates
are Running the Asylum" (which should be required reading before
a programmer is given a keyboard) and Alan Simpson's "dBase III
Programmers' Reference Manual"! And I buy a *lot* of books. :-)

-- 
Want to reply via email? Put the current month and year in 
my email address.



From reic0024 at bulldog5.d.umn.edu  Wed Nov  3 18:17:57 1999
From: reic0024 at bulldog5.d.umn.edu (Aaron J Reichow)
Date: Wed, 3 Nov 1999 17:17:57 -0600
Subject: Tkinter and Solaris 7 Sparc [Q]
In-Reply-To: <7vqblc$ham$1@nnrp1.deja.com>
References: <7vqblc$ham$1@nnrp1.deja.com>
Message-ID: <Pine.SO4.4.02.9911031716390.17271-100000@bulldog5.d.umn.edu>

> I just get an error saying that libX11.so.6.1 is needed. Is there a way
> to build tkinter without X11? Or I just need to download and install
> X11?

Don't you already have X11 installed, with Solaris (that would be that
which drives CDE/Solaris Desktop/OpenView)?  And if the answer is no, and
you're running simply from the console, why would you need Tkinter,
which provides a graphical interface?

Aaron




From dvbden at hotmail.com  Wed Nov 10 17:28:07 1999
From: dvbden at hotmail.com (dvbden at hotmail.com)
Date: 10 Nov 1999 17:28:07 -0500
Subject: $$ Information Auction - Sell What You Know -Buy What You Dont $$  9322
Message-ID: <3829f177$1@news.iglou.com>

Hi All,		
Check this site out - http://www.KnowToday.com

I recently registered on an interesting new site called - KnowToday.com that allows its members to buy and sell information! They are currently recruiting new members and will launch the site in several months. For
a limited time, they are offering an incentive plan, where they pay you
for successful referrals.

It sounds really interesting. You can go on their site and ask detailed questions or make money by answering questions. They are designing
the site with many different categories, so almost anyone will be able to ask and answer questions. Also, you can make money if any of your referrals use the site. 

It only takes a minute to join -  http://www.knowtoday.com/ -  they
explain everything on the web site. It's totally free and you have no 
further obligations after registering 

Hope you like it! 

John
zwgjopdryoeyzcxylsjccxoibdsjjclyvexodixixbobupssundesmltffbwmueyegmijdtbwstcyenqzdtwrjqrfyo




From gerrit.holl at pobox.com  Mon Nov  1 14:21:18 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Mon, 1 Nov 1999 20:21:18 +0100
Subject: __str__ vs. __repr__
In-Reply-To: <19991101121709.A134039715@vislab.epa.gov>
References: <19991101121709.A134039715@vislab.epa.gov>
Message-ID: <19991101202118.B29736@optiplex.palga.uucp>

Randall Hopper wrote:
> 
> 1) Is there a convention for what __str__ and __repr__ should return for
>    classes?

__str__ should return a string, __repr__ may also return, for example, a
dictionairy.

> 2) Or, whatever they return, should they return the same value?

I don't know: I think it's wrong to use both.

regards,
Gerrit.
-- 
linuxgames.nl.linux.org	 All about games on Linux in Dutch (under construction).
www.nl.linux.org	       The Dutch resource for Dutch Linux documentation.
www.nl.linux.org/discoverb		    Learn foreign words and definitions.
www.nl.linux.org/~gerrit/asperger	 I'm an Asperger, go here for more info.
www.nl.linux.org/~gerrit			    About me and pictures of me.



From pinard at iro.umontreal.ca  Tue Nov 30 11:56:55 1999
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 30 Nov 1999 11:56:55 -0500
Subject: '==' vs. 'is' behavior
In-Reply-To: "Fredrik Lundh"'s message of "Tue, 30 Nov 1999 09:03:09 +0100"
References: <382C2C6A.64A8F2C2@theriver.com> <14380.12036.197083.409565@weyr.cnri.reston.va.us> <8E7F86267duncanrcpcouk@news.rmplc.co.uk> <oqemd9dk2g.fsf@titan.progiciels-bpi.ca> <015301bf3b0a$685a3fc0$f29b12c2@secret.pythonware.com>
Message-ID: <oqg0xnapew.fsf@titan.progiciels-bpi.ca>

First of all, thanks for your replies, all.


"Fredrik Lundh" <fredrik at pythonware.com> ?crit:

> according to the eff-bot:

What is "the eff-bot"?  I've read about "eff-bot" and "tim-bot" a few
times by now, and thought it was some kind of internal joke in the group,
but begin to think there might be some substance behind the expression! :-)

>     foo is None
>     type(foo) is type(bar) # same type
>     foo is bar # same instance

> use '==' for everything else (unless you know what
> you're doing, of course).

Wise enough.  I guess I'll do exactly that.  Thanks for the suggestion.


"Tim Peters" <tim_one at email.msn.com> ?crit:

> [Fran?ois Pinard]
> > By the way, is there common wisdom (or rather, maybe,
> > usage-standards-to-be) about using `is' instead of `=='?

> In general, don't -- you'll often end up regretting this cleverness.

I've never been bitten so far, and as long as I do only sensible things,
there is no reason nor fear I would be bitten later.  I presume you agree
that Fredrik suggestions, above, are very reasonable?

[Many good explanations and examples not repeated.]

> Clearest and safest:  don't use "is" when "==" would suffice.

I'm a bit minimalist while programming.  Or, to say the same thing another
way, I rather like Arvo Part's music.  When a program source is just
abusing of parentheses without justification (helping `python-mode' _is_
a justification!), it gives me the uncomfortable feeling the author has
never been able to properly sort operator priorities, and probably many
other things as well :-).  Of course, I quite agree with you that there
is equilibrium and reason to put in the pondering.

And to be fully honest, I find the "is not" syntactical sugar rather sweet.
It would be a shame not using it, once in a while :-).


nascheme at enme.ucalgary.ca ?crit:


> Using `is' interchangably with == is asking for trouble.

Of course.  This is why I asked for some sounded advice about how to do
it nicely.

> Your lucky you don't do Lisp.

I don't? :-)

In fact, I like Scheme a lot, and use the Gambit system for some bigger
applications.  And, despite with much less pleasure, I sometimes play around
Emacs LISP.  If I could get Gambit and Python together, it would be fun!

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





From leonar_n at epita.fr  Mon Nov 29 09:32:25 1999
From: leonar_n at epita.fr (leon)
Date: Mon, 29 Nov 1999 15:32:25 +0100
Subject: Unbuffered stdout
References: <81u1k2$27qkk$1@oceanite.cybercable.fr> <046a01bf3a75$80b84d90$f29b12c2@secret.pythonware.com>
Message-ID: <81u2gm$26v0f$1@oceanite.cybercable.fr>

I use it into a cgi script, with apache, and for my IRC like, I must do
a thing like :
    while 1:
        strLine = self.pRoomFile.readline()
        while strLine:
            sys.stdout.write(strLine)
            sys.stdout.flush()
        strLine = self.pRoomFile.readline()
        time.sleep(1)

And the lines I print don't appear in the browser
Fredrik Lundh <fredrik at pythonware.com> wrote in message
news:046a01bf3a75$80b84d90$f29b12c2 at secret.pythonware.com...
> leon <leonar_n at epita.fr> wrote:
> > I'm developping an IRC like software which must work on the Web. I can't
use
> > Java, and I'm using Python to program the CGI. But, I have a big problem
: I
> > don't know how to make the stdout unbuffered.
> > I tried to make some
> >     sys.stdout.write("test")
> >     sys.stdout.flush()
> > but it don't work.
>
> looks fine to me.  maybe something else
> is wrong?
>
> anyway, you could try running the inter-
> preter with the -u option:
>
> $ python -?
> Unknown option: -?
> usage: python [option] ... [-c cmd | file | -] [arg] ...
> Options and arguments (and corresponding environment variables):
> ...
> -u     : unbuffered binary stdout and stderr (also PYTHONUNBUFFERED=x)
> ...
>
> </F>
>
>
>





From tbryan at python.net  Sat Nov  6 22:18:14 1999
From: tbryan at python.net (Thomas A. Bryan)
Date: Sat, 06 Nov 1999 22:18:14 -0500
Subject: making the Python case to my dept.
References: <9911051740130D.02530@william.jinsa.org> <3823FC43.CF7FBD71@python.net> <99110609085702.02970@quadra.teleo.net> <382467FA.577E70A0@callware.com>
Message-ID: <3824EF76.113E3699@python.net>

Ivan Van Laningham wrote:

> > Since this is a FAQ, I'd vote for preserving Thomas' answer somewhere
> > on the python.org Web site, so people could point to the URL when the
> > question comes up again.
> >
> 
> I agree.  His answer is nearly perfect.  Thanks, Thomas.

Thanks!  I'll rewrite some parts to make it more generally applicable 
and add it to the FAQ.  Do you think that section 2 (Python in the 
Real World) or section 1 (General information and availability) 
is most appropriate?

too-pleased-by-the-response-to-think-of-anything-witty-ly yours

---Tom



From jbaddor at sca.uqam.ca  Fri Nov 12 18:51:39 1999
From: jbaddor at sca.uqam.ca (Jean-Bernard ADDOR)
Date: Fri, 12 Nov 1999 18:51:39 -0500 (EST)
Subject: __doc__ print \012 on Debian Gnulinux2
Message-ID: <Pine.A41.4.05.9911121848040.37340-100000@maia.sca.uqam.ca>

Hey !

I try to make some documentation strings in my code but then I use them
the result is very difficult to read:

kili$  /usr/bin/python
Python 1.5.1 (#1, Dec 17 1998, 20:58:15)  [GCC 2.7.2.3] on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> xtGgrads.__doc__
'Apply x^t G grad sigma equation.\012\012    xtGgrads(G, (d0, d1),
(o0,o1)=(None,None), (s0, s1)=(1,1))\012    G GSI generator matrix\012
(d0, d1) sigma derivatives\012    (o0,o1) origin coordinate, center on
biggest side if unspecified\012    (s0, s1) distance between two pixel\012
'
How could I make it print return instead of \012?


	Jean-Bernard





From robin at alldunn.com  Tue Nov 16 00:13:33 1999
From: robin at alldunn.com (Robin Dunn)
Date: Mon, 15 Nov 1999 21:13:33 -0800
Subject: ANNOUNCE: wxPython 2.1.11
Message-ID: <007201bf2ff1$55d501b0$0301a8c0@easystreet.com>

wxPython 2.1.11 has been released and you can get your copy today from http://alldunn.com/wxPython/

New features in this release include HTML Printing, better support of GUI interaction from non-GUI threads, Clipboard, Drag and Drop, font enumerator, a new calendar widget, many little tweaks and fixes.  Please see http://alldunn.com/wxPython/README.txt for details.

--
Robin Dunn
Software Craftsman
robin at AllDunn.com
http://AllDunn.com/robin/
http://AllDunn.com/wxPython/  Check it out!






From usenet-mm-yyyy at julesallen.com  Fri Nov 19 06:55:58 1999
From: usenet-mm-yyyy at julesallen.com (Jules Allen)
Date: Fri, 19 Nov 1999 11:55:58 GMT
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <JlFX3.3659$s87.67930@dfiatx1-snr1.gtei.net> <80oqn9$ocq$2@newshost.accu.uu.nl> <oLUX3.286$yA4.4821@dfiatx1-snr1.gtei.net> <80p9su$ppo$2@newshost.accu.uu.nl> <80piru$13rl$1@hub.org> <80re5d$55a$1@newshost.accu.uu.nl> <fjPY3.770$An3.14793@newsfeed.avtel.net> <810jto$lo0$1@newshost.accu.uu.nl> <811pdu$dvo$1@newshost.accu.uu.nl>
Message-ID: <iVaZ3.56$AZ6.2217@dfiatx1-snr1.gtei.net>

In article <811pdu$dvo$1 at newshost.accu.uu.nl>, Martijn Faassen wrote:
| > No, no, a more extensive deja search indicated to me that this person has
| > posted under other names to the group previously. At least, unless there are
| > 10 people with the same speech patterns posting to comp.lang.python. :)
| 
| with the same speech patterns _from vex.net_, I meant! deja nor I are that
| good at scanning for similar speech patterns yet, but the vex.net subset
| is doable for the eyeball. 

Tim assures me this is purely an email address anti-spam obfuscation
method and not a witness protection plan.

My-anti-spam-methods-are-positively-transparent-in-comparison-ly-yrs
Jules

-- 
Want to reply via email? Put the current month and year in 
my email address.



From gawron at obop.com.pl  Fri Nov 12 10:56:32 1999
From: gawron at obop.com.pl (=?iso-8859-2?Q?Przemys=B3aw?= G. =?iso-8859-2?Q?Gawro=F1ski?=)
Date: Fri, 12 Nov 1999 16:56:32 +0100
Subject: DBF interface
Message-ID: <382C38AF.69FFACF4@obop.com.pl>

Is there a DBF file interface module ?

Thanks

Przemek

--
Przemyslaw G. Gawronski UIN:8358522
mailto:gawronskip at usa.net
mailto:gawron at obop.com.pl





From gerrit.holl at pobox.com  Tue Nov 16 11:31:42 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Tue, 16 Nov 1999 17:31:42 +0100
Subject: Beginning programming with Python
In-Reply-To: <antigorite-80rol2/INN-2.2.1/attentive@broadway.news.is-europe.net>
References: <sng0ychgwt.fsf@motorola.com> <382B3A53.5816B2@callware.com> <382B3D79.74CD3E0F@email.sps.mot.com> <80p9ds$ppo$1@newshost.accu.uu.nl> <asteroidal-80psfo/INN-2.2.1/bat@broadway.news.is-europe.net> <38312D55.A36EB56D@iqsoft.hu> <antigorite-80rol2/INN-2.2.1/attentive@broadway.news.is-europe.net>
Message-ID: <19991116173142.A31379@optiplex.palga.uucp>

> You could take the number of man months necessary or you
> could try to quantify the quality of you product.

Yes... Those two are always "omgekeerd evenredig". Can any of the
Dutch-speaking guys in here translate it for the not-Dutch-speaking
people? :-)

But serious:
the less time it costs to program it, the more time it costs to run it,
the more time it costs to program it, the less time it costs to run it.

How long would it take to build a minimalistic web server in c? More then
1 Line Of Code (LOC), I think :-) And I think it'd run much, much faster
than the python-import version...


how-many-databases-are-written-in-sh-with-dd-only?,
-- 
We are using Linux daily to UP our productivity - so UP yours!
(Adapted from Pat Paulsen by Joe Sloan)



From tiddlerdeja at my-deja.com  Tue Nov 23 14:10:50 1999
From: tiddlerdeja at my-deja.com (tiddlerdeja at my-deja.com)
Date: Tue, 23 Nov 1999 19:10:50 GMT
Subject: dir() and propertly list NEWBIE
Message-ID: <81eorl$b60$1@nnrp1.deja.com>

Is it possible to get the functions of an object using dir()? Rather
that looking them up in a book/API.

Similarly, If I can get an objects functions, is it possible then get a
list of the possible parameters for that function?

Any help appreciated.


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



From tim_one at email.msn.com  Wed Nov 10 01:39:13 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 10 Nov 1999 01:39:13 -0500
Subject: __str__ vs. __repr__
In-Reply-To: <14376.31199.948463.211294@goon.cnri.reston.va.us>
Message-ID: <000301bf2b46$4d7e6f60$d82d153f@tim>

[Toby Dickenson:]
> Tim's Rat class is a good example of how __repr__ should not be
> implemented, because it exposes the internal representation of
> the object, not the external value.

[Rat's Tim class]
> Beg to differ:  it's an excellent example of how __repr__ should
> be implemented.

[Jeremy Hylton]
> It's no accident that the function is called __repr__.  It's role is
> to expose the representation.

No, in that case it would have been named __representation__.  __repr__ is
actually an acronym, meaning Regular-Expression Parsable Rendition.
Similarly __str__ is short for Simpler Than Repr.  I never learned what
__init__ stands for -- it never interested Tim <wink>.

guess-what-wink-stands-for-and-win-a-fabulous-prize!-ly y'rs  - tim





From ivanlan at callware.com  Wed Nov 10 14:01:40 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Wed, 10 Nov 1999 12:01:40 -0700
Subject: Tkinter PhotoImage
Message-ID: <3829C114.77389F49@callware.com>

Hi All--
Ok, I've managed to create an image, using xx=PhotoImage(file), and I
can retrieve pixel values from that image using xx.get(x,y).

Now I want to use the put() method to write a pixel.

How do I do that?  I would have expected that put() would take a string
of pixel values for the data argument, but that doesn't seem to be the
case. ...

And what's the bbox argument?

<there's-a-cat-on-my-screen>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours
----------------------------------------------



From alan-ng at empresa.com  Mon Nov  1 07:48:11 1999
From: alan-ng at empresa.com (Alan Pogrebinschi)
Date: Mon, 1 Nov 1999 07:48:11 -0500
Subject: Help! Can't install MySQLdb
Message-ID: <nXfT3.78$f16.2191@news2.mia>

Hi!

I am new to python modules, I am triying to install MySQLdb 0.1 on RH 5.2 ,
Python 1.5.2, and get (when running "build.py" as root) :

cp: /usr/lib/python1.5/config/Makefile.pre.in: No such file or directory
make: Makefile.pre.in: No such file or directory
make: *** No rule to make target `Makefile.pre.in'.  Stop.
make: *** No targets.  Stop.
Traceback (innermost last):
  File "build.py", line 14, in ?
    import MySQLdb
  File "MySQLdb.py", line 19, in ?
    from _mysql import *
ImportError: No module named _mysql


It is strange that the dir "/usr/lib/python1.5/config" doesn't even exist
(the parent exists). I used the RPM to upgrade Python to 1.5.2 (I had
1.5.1).

I tried to install the older mysql driver too (the one that comes with
ZMySQLDA) and got a very similar error msg when trying to "make" it.

Any hint??

Thanks.

Alan





From gerrit.holl at pobox.com  Mon Nov  1 09:06:00 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Mon, 1 Nov 1999 15:06:00 +0100
Subject: Newbie Question
In-Reply-To: <7vidsi$qm8$1@nnrp1.deja.com>
References: <7vidsi$qm8$1@nnrp1.deja.com>
Message-ID: <19991101150600.B26174@optiplex.palga.uucp>

strangiato at my-deja.com wrote:
> I am completely new to the language.
> 
> I heard that Yahoo uses it?
> Is this true?

I don't know.

> Is it a good language for developing Web Pages?

Yes.

> Is it a "best of breed"?

Yes again.

regards,
Gerrit.
-- 
linuxgames.nl.linux.org	 All about games on Linux in Dutch (under construction).
www.nl.linux.org	       The Dutch resource for Dutch Linux documentation.
www.nl.linux.org/discoverb		    Learn foreign words and definitions.
www.nl.linux.org/~gerrit/asperger	 I'm an Asperger, go here for more info.
www.nl.linux.org/~gerrit			    About me and pictures of me.



From greg.ewing at compaq.com  Wed Nov 17 05:54:00 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Wed, 17 Nov 1999 23:54:00 +1300
Subject: Type/Class Unification in Ruby
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp> <7vjtrc$hmm$1@newshost.accu.uu.nl> <38204614.2EAC6051@compaq.com> <5ld7trtak5.fsf@eric.cnri.reston.va.us> <3826D63C.2731BB1F@compaq.com>
Message-ID: <38328948.9E35556F@compaq.com>

I've been looking into the way types and classes are
unified in Ruby.

As I mentioned earlier, all objects in Ruby are associated
with an object of class Class, which plays the role of
both the PyTypeObject and PyClassObject. Consequently,
it is possible to define subclasses of any of the built
in types and add or override methods. It's even possible
to add instance variables -- I'll explain how that works
below.

Each of the built-in types is represented by a unique
C structure, as in Python. Therefore it is important that
when a subclass of a built-in class is instantiated, the
right kind of C structure is created, since all the
built-in methods of its superclass expect to get that
kind of structure.

This is guaranteed by two characteristics of Ruby that
differ from Python:

1) Instantiation of a class is done by calling a class
method of that class. (Ruby has real class methods and
variables, like Smalltalk. In fact, Ruby's internals are
very Smalltalk-like in many ways.)

2) A class can only inherit from at most one other
class.

The only methods which actually create C structures are
the built-in instantiation methods of the built-in
classes, so even if you define your own instantiation
method in your subclass, it eventually has to call a
built-in one somewhere up its inheritance chain, which
knows what kind of C structure is needed.

This scheme will fail if you somehow manage to extract
a built-in method out of one class and stuff it into
another. However there doesn't seem to be any obvious
way to do this (nothing like myClass.__dict__["foo"] in 
Python). So it seems to be safe.

Now, what happens if you assign to a new instance variable
in a method of your subclass of a built-in class? At first
sight it seems that this can't possibly work, because the
C structs of built-in types don't have any place to store
arbitrary instance variables.

When I tried it, however, to my surprise, it worked!
It turns out that there is a global table whose purpose
is to hold the values of these instance variables (in
Python terms it is a dictionary of dictionaries, indexed
by [object][variable_name]). Python might be able to
use a similar solution, although it might cause ref
counting problems.

As a final note, some built-in classes, such as Fixnum,
don't have any instantiation method, so although you can
subclass them, there's no point because you can never
create an instance of your subclass.

Others, however, such as Array (Ruby's version of ListType)
do have an instantiation method, and subclasses work
just the way you'd expect.

Hoping this has been enlightening,
Greg



From gmcm at hypernet.com  Sat Nov  6 20:58:57 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sat, 6 Nov 1999 20:58:57 -0500
Subject: Job skills [was Re: making the Python case to my dept.]
In-Reply-To: <m3n1srnoun.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <1270182561-5943751@hypernet.com>

[Michael Hudson]
> > > What are mathematicians' critical job skills? I'm curious,
> > > being one, and not having being taught anything resembling a
> > > `critical job skill' yet, and happy about it.

[me]
> > The ability to work long and hard on entirely meaningless
> > problems.
> > 
[Michael]
> Fair comment.

Maybe you don't realize that I was a math major too.

and-good-preparation-it's-been-ly y'rs

- Gordon



From jhauser at ifm.uni-kiel.de  Thu Nov 18 08:54:47 1999
From: jhauser at ifm.uni-kiel.de (Janko Hauser)
Date: 18 Nov 1999 14:54:47 +0100
Subject: Comparing Python to (free) Math languages/environments
References: <19991118014716.13748.00000334@ng-bj1.aol.com> <810rev$jsm$1@nnrp1.deja.com>
Message-ID: <874sej7vmg.fsf@ifm.uni-kiel.de>

siegfried.gonzi at kfunigraz.ac.at writes:

> No. I also, tried to get into Python. But it is too "administrative" and
> complicated. Than I came to Yorick. Yorick is ranked as a very good
> substitute for IDL and also Matlab. It is or better it should be very
> fast (compared to Matlab, Scilab or RLab or Octave). And it has stunnig
> graphic features (default).
> 
> What Yorick is lacking of is GUIs. This than, is a matter of taste. But
> for scientists, which only look for good code, should this be not an
> obstacle.

This is exactly the strength of this approach, to extend a general
scripting language with numerical capabilities. The connection between
other worlds, be it GUI, Databases, WEB, serialIO, OpenGL, fileformats
... is much cleaner and easier in Python than in all the other
systems. 

And then look how things like OO are implemented in Matlab or higher
rank arrays. It is right that the syntax is more verbose in NumPy than
in Matlab, but then this appeals to some as a good feature, too :-)

In the end I want to add, that NumPy does connect to the same graphics
engine like Yorick (Gist), so it should be possible to achieve the
same output.

__Janko
-- 
  Institut fuer Meereskunde             phone: 49-431-597 3989
  Dept. Theoretical Oceanography        fax  : 49-431-565876
  Duesternbrooker Weg 20                email: jhauser at ifm.uni-kiel.de
  24105 Kiel, Germany



From marc at oscar.noc.cv.net  Wed Nov  3 03:19:58 1999
From: marc at oscar.noc.cv.net (User &)
Date: Wed, 03 Nov 1999 08:19:58 GMT
Subject: Need Berkley DB!!
References: <381D5E84.98931682@bibsyst.no> <hAQT3.488$bu3.39846291@news.randori.com> <381FE1D5.C0F721A6@bibsyst.no>
Message-ID: <slrn81vrpb.1hjh.marc@oscar.noc.cv.net>

In article <381FE1D5.C0F721A6 at bibsyst.no>, Thomas Weholt wrote:
>Robin Dunn wrote:
>
>> > After I spent most of my weekend trying to get Berkley DB to work with
>> > Python on my Linux ( Suse 6.1) machine, I`ve now given up completly.
>> > I`ve tried to install DB.1.85. and DB.2.7.7, neither worked, either
>> > crashing under installation ( which of course, is not Python
>> > related ) or when I tried to test the installation with the examples in
>> > the documentation ( the wrapper for the 2.7.5-module listed at
>> > www.python.org, under modules, databases ).
>>
>> Have you tried my DB module at http://starship.python.net/crew/robind/ ?
>> Assuming you can get DB 2.7.x built and installed, this module should do
>> what you need.
>>
>
>Yes, I`ve tried it and the installation did go smooth, but when I tried to
>use the example in the documentation Python barfed some errors at me, then
>died with a scream. I can probably recreate the errors and send you a copy
>of the output.
>
>I think the problem might be solved if I, sigh, re-install Red Hat 6.1 and
>trash that oversized, over-featured Suse 6.2 Linux distro I got now. The Red
>Hat version of Python came with all that stuff installed. There`s not enough
>hours in a day to keep working on things like this. At least not right now.
>
>Thanks for your interest though.
>
>Best regards,
>Thomas Weholt
>

off topic, but instead of linux you might want to try FreeBSD,
It is much less agravation, form my personal experence anyway.

Good Luck

Marc



From oofoe at my-deja.com  Wed Nov  3 18:30:19 1999
From: oofoe at my-deja.com (oofoe at my-deja.com)
Date: Wed, 03 Nov 1999 23:30:19 GMT
Subject: Namespace question...
Message-ID: <7vqgib$l8v$1@nnrp1.deja.com>

Hi!

How can you access the global namespace in python 1.5.1? I'd like to get
a version string which is declared in the main script file. However,
each module also has a version string which overrides the original one.
I tried to access it as __main__.Version, similarly to $main::Version in
Perl, which is what I'm more familiar with. Of course, it didn't work.

I have something like this:

module file foo.py:
------------------------------
Version = '$Id$'

def bar():
   print "Version is: ", __main__.Version

main program:
------------------------------
Version = '$Id$'

import foo

foo.bar()

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

I'd certainly appreciate any pointers. The namespace stuff seems to get
a little hazy around issues like this.

Thanks!

Jos'h


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



From boud at rempt.xs4all.nl  Fri Nov 26 15:54:47 1999
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 26 Nov 1999 20:54:47 GMT
Subject: XPython (X11)
References: <383E5350.1E07536A@xrce.xerox.com> <slrn83tj53.k7h.bernhard@alpha1.csd.uwm.edu>
Message-ID: <81ms2n$126$1@essle.valdyas.org>

Bernhard Reiter <bernhard at alpha1.csd.uwm.edu> wrote:
> On Fri, 26 Nov 1999 10:30:56 +0100, Emmanuel Pietriga <emmanuel.pietriga at xrce.xerox.com> wrote:
>>Is anyone using XPython around here?

> This is the first time I heard of XPython. :)
> The only reference I found is :
> 	http://ftp.uni-klu.ac.at/programming/python/X-extension/Top.html

>>I need help, or at least links to examples (I haven't found any but the
>>ones within the package)

> If I want to program a GUI I look in to wxPython.
> Or pyGtk or pyFOX or pyFLTK, but XPython... :)

Hey! Don't forget PyQt and PyKDE (and PyQt seems to be
working under Windows, too, nowadays)...

> 	Bernhard

-- 

Boudewijn Rempt  | http://denden.conlang.org



From ivnowa at hvision.nl  Mon Nov  8 12:27:27 1999
From: ivnowa at hvision.nl (Hans Nowak)
Date: Mon, 8 Nov 1999 18:27:27 +0100
Subject: Popup menus in Tkinter
In-Reply-To: <806rtk$4e6$1@nnrp1.deja.com>
Message-ID: <199911081726.SAA25412@axil.hvision.nl>

On 8 Nov 99, dkaznadzey at my-deja.com wrote:

> Does anyone know weather there is a way to create
> a popup menu in tk (using Tkinter)? It seems
> like in order to be shown the menu must
> grow from the valid menubar. If there is a way
> to display standalone menu pane at given
> coordinates, please let me know.

Try this code... You can right-click on the canvas to make the menu 
pop up.

# menutest.py
# Makes a window pop up by rightclicking the canvas.

from Tkinter import *
import regsub

def getgeometry(something):
    s = something.geometry()
    return map(int, regsub.split(s, "[x+]"))

root = Tk()

root.canvas = Canvas(root, height=100, width=100,   
background='#674689')
root.canvas.pack()

def new_file(event=None):
    print "Opening new file"

menu = Menu(root)
menu.add_command(label="New...", underline=0, command=new_file)
menu.add_command(label="Open...", underline=0, command=new_file)
menu['tearoff'] = 0

def activate_menu(event=None):
    sizex, sizey, x, y = getgeometry(root)    
    menu.tk_popup(x+event.x, y+event.y)

root.canvas.bind("<Button-3>", activate_menu)

root.mainloop()

--Hans Nowak (zephyrfalcon at hvision.nl)
Homepage: http://fly.to/zephyrfalcon
You call me a masterless man. You are wrong. I am my own master.



From tim_one at email.msn.com  Wed Nov 24 16:13:49 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 24 Nov 1999 16:13:49 -0500
Subject: Python complaints
In-Reply-To: <1268672881-3462615@hypernet.com>
Message-ID: <000701bf36c0$cdb9f480$292d153f@tim>

[Thomas A. Bryan]
> ...
> For example, I saw some code that I wanted to copy in DejaNews
> c.l.py archives, but the lines wrap in narrow columns in DejaNews.
> I had to reindent the code.  Grrr.

In DejaNews, while looking at a msg, click the "View original Usenet format"
link following the msg end:

    Track this thread for me

    Unsubscribe from comp.lang.python
    Mail this message to a friend
    View original Usenet format   ************* THIS ONE **************
    Create a custom link to this message from your own Web site

> I have also had problems when e-mailing code when I e-mail the
> code as text in the message instead of attaching a document.
>
> Is there a better way to transmit code so that possible
> indentation munging won't force the reader to reindent?

If you don't want to use attachments (chicken <wink>), avoid all control and
high-bit characters other than newline (which means, in particular, don't
use tabs!  use spaces), and keep your maximum line length less than 80
characters.

good-advice-for-mailing-programs-in-any-language!-ly y'rs  - tim





From fredrik at pythonware.com  Wed Nov 24 09:06:03 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 24 Nov 1999 15:06:03 +0100
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <867lj85g4c.fsf@g.local> <al8yabo9fm3.fsf@sirppi.helsinki.fi>
Message-ID: <00bf01bf3685$0c345610$f29b12c2@secret.pythonware.com>

Markus Stenberg <mstenber at cc.Helsinki.FI> wrote:
> >       - a less yucky way of saying "for i in range(1,20,3)"
> 
> see above - although some special syntactic sugar would be "nice", I prefer
> minimalistic approach over say, C.. of course, Dylan does it (IMNSHO)
> cleanest:
> 
> for (i from 3 to 19 by 3)
> ...

note that python already has a special syntax
for this purpose:

>>> a = range(20)
>>> print a[1:20:3]
Traceback (innermost last):
  File "<stdin>", line 1, in ?
TypeError: sequence index must be integer

(note that you got a type error, not a syntax
error.  numpy arrays fully support this syntax).

in other words, you don't have to be guido to
realize that something like:

    for i in [1:20:3]:
        ...

could be the right way to do it in Python.  in
fact, it's already on the list for 2.0:

http://www.foretec.com/python/workshops/1998-11/proceedings/guido/sld011.htm

</F>




From ivanlan at callware.com  Wed Nov  3 23:54:07 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Wed, 03 Nov 1999 21:54:07 -0700
Subject: Python / Perl joke in rec.humor.funny
References: <38210E22.BBDAF855@erols.com>
Message-ID: <3821116F.113DFE75@callware.com>

Hi All--

"Edward C. Jones" wrote:
> 
> A Python vs. Perl joke has appeared in the moderated,
> low-traffic, humor group rec.humor.funny. It can be found at
> 
> http://www.netfunny.com/rhf/jokes/99/Nov/perl.html
> 

Pretty good, and true, except that that last line out to read "six
minutes from now" instead of six months. ...

<revenge-of-the-flea-dirt>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From jjkunce at my-deja.com  Mon Nov  1 11:13:01 1999
From: jjkunce at my-deja.com (jjkunce at my-deja.com)
Date: Mon, 01 Nov 1999 16:13:01 GMT
Subject: How to control app from Python?
References: <38160DD8.5C695E09@theriver.com> <3817BD5F.7E05BDC9@home.com> <38185A03.9AB0FAD6@theriver.com>
Message-ID: <7vke68$6r6$1@nnrp1.deja.com>

In article <38185A03.9AB0FAD6 at theriver.com>,
  Andy Fox <foxes at theriver.com> wrote:
> The app to be controlled is visual...
> I discovered Mark Hammond's win32 stuff...

Can the app be controlled via COM? If so, be sure to look into Mark's
win32com stuff. I have some examples of using COM to control Inernet
Explorer and GroupWise, if you are interested.

  --Jeff


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



From fdrake at acm.org  Thu Nov 11 11:37:36 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Thu, 11 Nov 1999 11:37:36 -0500 (EST)
Subject: docs
In-Reply-To: <80epm7$vbo$1@nnrp1.deja.com>
References: <80epm7$vbo$1@nnrp1.deja.com>
Message-ID: <14378.61648.633649.431232@weyr.cnri.reston.va.us>

mehta at mama.indstate.edu writes:
 > I have noticed for the local docs of python in IE5 on winnt pages
 > randomly donot get rendered but are left as raw html in browser as I go
 > back and forth using browser buttons.  has any one experienced same

  I have never seen this; if anyone can reproduce this and help
isolate the cause I'd appreciate hearing about it.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From mcfletch at vrtelecom.com  Thu Nov 11 16:27:32 1999
From: mcfletch at vrtelecom.com (Mike Fletcher)
Date: Thu, 11 Nov 1999 16:27:32 -0500
Subject: ASCII delimited files
In-Reply-To: <YMDW3.5675$JD1.379317@typhoon.nyroc.rr.com>
Message-ID: <NDBBIAPHEKGOBIHCBLIIIEOICAAA.mcfletch@vrtelecom.com>

Okay, here's an entirely different approach, extremely memory intensive,
fairly fast.  Gives pretty good speed up to the point where you start
hitting memory effects (my 256 MB PIII 500 loads 10MB in 45 seconds, but
takes 7.5 minutes for a 36MB file).

For even greater speed, using an index-counting scheme (instead of expensive
front-of-list-deletions) might be tried.

Note: this does none of the fancy delimiter-determination of delimited.py,
it's just a loading engine.

Cheers,
Mike

Time for file of 365 bytes
 (5 * testset, 11 records found):
 0.0
Time for file of 3650 bytes
 (50 * testset, 101 records found):
 0.00999999046326
Time for file of 36500 bytes
 (500 * testset, 1001 records found):
 0.120000004768
Time for file of 365000 bytes
 (5000 * testset, 10001 records found):
 1.25199997425
next test may take 15 seconds or so
Time for file of 3650000 bytes
 (50000 * testset, 100001 records found):
 13.5299999714
next test may take 45 seconds or so
Time for file of 10950000 bytes
 (150000 * testset, 300001 records found):
 50.4729999304


-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Darrell
Sent: November 11, 1999 1:20 PM
To: python-list at python.org
Subject: Re: ASCII delimited files


Oops.
That fix pushed into or beyond exponential runtime.
It won't finish the 3.6meg file anytime soon.
String slicing and pasting is dirt slow.

E:\ace\dtd\q\delim1>delimited.py
3600 bytes
0.0310000181198 time

E:\ace\dtd\q\delim1>delimited.py
36000 bytes
3.04700005054 time

Got some improvement with the attached version at
the cost of memory.

Now I'm thinking about the 'C' version again.

E:\ace\dtd\q\delim1>delimited.py
3600
0.0310000181198

E:\ace\dtd\q\delim1>delimited.py
36000
0.327999949455

E:\ace\dtd\q\delim1>delimited.py
360000
3.21899998188

E:\ace\dtd\q\delim1>delimited.py
3600000
48.2030000687

--
--Darrell




--
http://www.python.org/mailman/listinfo/python-list


-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: csv.py
URL: <http://mail.python.org/pipermail/python-list/attachments/19991111/78894619/attachment.ksh>

From billtut at microsoft.com  Wed Nov 10 13:25:10 1999
From: billtut at microsoft.com (Bill Tutt)
Date: Wed, 10 Nov 1999 10:25:10 -0800
Subject: Python company worth $100M
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8100DB9151E@RED-MSG-50>


> From: Mike Fletcher [mailto:mcfletch at vrtelecom.com]
> 
> 
> Nope, don't think you're missing anything. If I recall correctly, the
> asyncore docs even mention their assumption that the 
> processing to be done
> per call is trivial.  Since all other channels are "blocked" while the
> compute-heavy call is processing, the system will suffer performance
> degradation for all other clients.  Thread-pools seem to be a decent
> approach for longer-running calls.
> 

Only if they're synchronous operations, the really crazy people move
completly to an asynchronous finite state machine, have one thread per CPU
(and that thread only runs on that CPU), and schedule their work in chunks
so that one hunk of code can process multiple "work packets" at once,
thereby playing nice with the memory caches. There are apps that need this
performance and actually do this btw. Not sure if you can schedule a thread
to sit on one CPU on anything besides NT though. I'd be interested to know
otherwise...

Bill



From lany at iname.com  Tue Nov 16 18:47:35 1999
From: lany at iname.com (Lany Wilson)
Date: Tue, 16 Nov 1999 18:47:35 -0500
Subject: Can't build python1.5.2 on Slackware
Message-ID: <3831ED17.5283BF03@iname.com>

The compile bombs on socketmodule.c. Something about the way gethostbyname_r is
defined. Does anyone know of a quick fix? I'm using Slackware 4/egcs.

Tia-
L Wilson



From thomas at bibsyst.no  Tue Nov  2 08:30:41 1999
From: thomas at bibsyst.no (Thomas Weholt)
Date: Tue, 02 Nov 1999 14:30:41 +0100
Subject: General q. `bout subclassing
Message-ID: <381EE780.EBF18F59@bibsyst.no>

Hi,

Simple one :

How do I subclass fundamental stuff like lists etc,

tried something like :

> >>> class my_list([]):
> ...     a = ''
> ...
> Traceback (innermost last):
>   File "", line 1, in ?
> TypeError: base is not a class object
> >>>
>

Huh ??

Thomas Weholt




From dave.rose at wcom.com  Tue Nov 16 18:41:55 1999
From: dave.rose at wcom.com (Dave Rose)
Date: Tue, 16 Nov 1999 23:41:55 GMT
Subject: Unix Message Queues
Message-ID: <7ZlY3.7116$dz1.594280@pm02news>

Has anyone been able to use unix message queues with Python.  I have been
trying to get them to work for a couple of days now and have been
unsuccessful.  I wrote an extension module in 'C' to complete the actually
message manipulation and the main function is a python thread.  The problem
I am having is that when I put the two together, it no longer functions as a
thread, you go into the function and never come out.  The thread is not
spawned.  Any help would be great.  The code is attached for the python
portion.


Thanks
Dave Rose





begin 666 test_proc.py
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M8V5S<R!W:6QL('-C<F5E;B!T:&4 at 86QA<FT at 86YD('1H96X at 8F%S960@(",-
M"B,@;VX@=&AE(&%L87)M(&YU;6)E<B!A;F0@=&AE(&-O;F9I9W5R871I;VX@
M:6YF;W)M871I;VX@=&AA="!I="!H87,N($ET('=I;&P@(" C#0HC(&9O<G=A
M<F0@=&AE(&%L87)M(&]N('1O('1H92!A;GD at 8V]N;F5C=&5D("=&4D5$)R!P
M<F]C97-S97,N(%1H92!A;&%R;2!W:6QL(" @(PT*(R!B92!L;V=G960@=&\@
M=&AE(&1A=&%B87-E(&%S('1H870 at 9G5N8W1I;VYA;&ET>2!I<R!D969I;F5D
M+B @(" @(" @(" @(" @(" @(",-"B, at 5&AI<R!P<F]C97-S('=I;&P at 8F4@
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M<F]M('-Y<R!I;7!O<G0 at 97AI= T*9G)O;2!T:')E861I;F<@:6UP;W)T("H-
M"F9R;VT at 475E=64@:6UP;W)T("H-"@T*:6UP;W)T('!Y7W!M#0II;7!O<G0@
M<W1R:6YG( T*9G)O;2!T:6UE(&EM<&]R=" J#0II;7!O<G0@;W,-"@T*9G)O
M;2!!;&%R;5]-<V<@:6UP;W)T("H-"F9R;VT at 9W5I8V]M;6]N(&EM<&]R=" J
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M(" @(" @(" @(" @(" @(" @(",-"B, at 86QA<FT@<V5R=FEC92XN+B @(" @
M(" @(" @(" @(" @(" @(" @(" @(" @(" @(" @(" @(" @(" @(" @(" @
M(" @(" @(" @(" C#0HC(",C(R,C(R,C(R,C(R,C(R,C(R,C(R,C(R,C(R,C
M(R,C(R,C(R,C(R,C(R,C(R,C(R,C(R,C(R,C(R,C(R,C(R,C(R,C(R,C(R,C
M(R,@(PT*(TQ)4U1%3E-%4E8],0T*(T%,05)-4T525CTR#0HC15A)1E-%4E8]
M,PT*(U-#4T525CTT#0HC3U-315)6/34-"B-&4D5$4T525CTV#0HC3TU315)6
M/3<-"B-#2$%.4T525CTX#0H-"DQ)4U1%3E-%4E9&4D]-/3$-"D%,05)-4T52
M5D923TT], at T*15A)1E-%4E9&4D]-/3,-"E-#4T525D923TT]- T*3U-315)6
M1E)/33TU#0I&4D5$4T525D923TT]- at T*3TU315)61E)/33TW#0H-"F-L87-S
M(%!R;V-397)V*%1H<F5A9"DZ#0H)9&5F(%]?:6YI=%]?*'-E;&8L9G)E860L
M9G=R:71E+&EN<75E=64L9W)O=7 ]3F]N92P@=&%R9V5T/4YO;F4L(&YA;64]
M3F]N92P-"@D)87)G<STH*2P@:W=A<F=S/7M]+"!V97)B;W-E/4YO;F4I. at T*
M"0E4:')E860N7U]I;FET7U\H<V5L9BP at 9W)O=7 L('1A<F=E="P@;F%M92QA
M<F=S+"!K=V%R9W,L('9E<F)O<V4I#0H)"6EN;F%M92 ]("=0<F]C4V5R=B<-
M"@D)<')I;G0@(DEN(%!R;V-M86X at 26YI="(-"@D)<V5L9BYS971.86UE*&EN
M;F%M92D-"@D)<V5L9BY4;F%M93US96QF+F=E=$YA;64H*2 -"@D)<V5L9BYQ
M=65U92 ](&EN<75E=64-"@D)<V5L9BYP=W)I=&4@/2!F=W)I=&4-"@D)<V5L
M9BYP<F5A9" ](&9R96%D#0H)"7-E;&8N4')O8V5S<TAD<B ]($YO;F4-"@D)
M<V5L9BY0<F]C97-S1'1L(#T at 3F]N90T*#0H)9&5F(')U;BAS96QF*3H-"@D)
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M<F]C;6%N44ED(#T at 4')O8VUA;E%)9 T*"0ES96QF+D9R9611260@/2!&<F5D
M44ED#0H)"6EF(%!R;V-M86Y1260@/" P. at T*"0D)<')I;G0@(D5R<F]R($-R
M96%T:6YG('1H92!1=65U97,B#0H)"65L<V4Z#0H)"0DC('=E(&YE960@=&\@
M<V5N9"!A(')E<75E<W0@=&\@<')O8VUA;B!F;W(@=&AE('!R;V-E<W,@:&5A
M9&5R(&EN9F\-"@D)"5!Y4&TN<')O8VAD<BA0<F]C;6%N44ED+$9R9611260I
M#0H)"0DC('1H:7,@=VEL;"!S96YD(&$@<F5Q=65S="!T;R!P<F]C;6%N(&9O
M<B!P<F]C97-S(&AE861E<@T*"0D)(R!W92!N965D('1O(&-A;&P at 8VAE8VMQ
M('1O(&-H96-K(&9O<B!T:&4@<F5P;'DN($EF( T*"0D)(R!W92!R96-E:79E
M(&$@<F5P;'D@=&AA="!W92!A<F4@;F]T(&5X<&5C=&EN9R!T:&5N#0H)"0DC
M('=E('=I;&P@:&%V92!T;R!H86YD;&4@:70N#0H)"0DC('=E(&%R92!S96YD
M:6YG(&$@+3$@87,@=&AE(%-C:7 @<75E=64@<V\@=&AA="!T:&4@<V-I<"!Q
M=65U92 -"@D)"2, at 9&]E<R!N;W0@:6YT97)U<'0@=&AE(&-A;&P at 9F]R($AE
M861E<B!A;F0 at 1&5T86EL($EN9F]R;6%T:6]N( T*#0H)"0EC:&5C:U]R=')N
M(#T at 4'E0;2YC:&5C:W$H4')O8VUA;E%)9"Q&<F5D44ED+"TQ+%!23T--04Y?
M5$E-14]55"D-"@D)"2-C:&5C:U]R=')N/2 at Y+# L,"D-"@D)"6EF('1Y<&4H
M8VAE8VM?<G1R;BD@/3T at 5'5P;&54>7!E. at T*"0D)"2,@;F5E9"!T;R!C:&5C
M:R!;,%T at 9F]R(&UE<W-A9V4@='EP90T*"0D)"6UE<W-A9V5?='EP92 ](&-H
M96-K7W)T<FY;,%T-"@D)"0EI9B!M97-S86=E7W1Y<&4@/3T at .3H-"@D)"0D)
M(R!P<F]C965D('1O(&-O9&4@=&\@<F5A9"!T:&4@<V-I<"!P:7!E#0H)"0D)
M"4YO;F4-"@D)"0EE;&EF(&UE<W-A9V5?='EP92 ]/2 P. at T*"0D)"0DC('=E
M(&AA9"!A('1I;65O=70 at 8V]N9&ET:6]N(&YE960-"@D)"0D)(R!I;B!T:&ES
M(&-A<V4@=&AI<R!I<R!A;B!E<G)O<B!A;F0@=V4@;F5E9"!T;R!R97!O<G0@
M:70N#0H)"0D)"7!R:6YT(")5;F%B;&4@=&\@:6YI=&EA;&EZ92!P<F]C;6%N
M(# B#0H)"0D)96QI9B!M97-S86=E7W1Y<&4@/3T@,3H-"@D)"0D)4')O8V5S
M<TAE861E<B ](&-H96-K7W)T<FY;,5T-"@D)"0D)4')O8V5S<T1T;" ](&-H
M96-K7W)T<FY;,ET-"@D)"0D)<V5L9BY0<F]C97-S2&1R(#T at 4')O8V5S<TAE
M861E<@T*"0D)"0ES96QF+E!R;V-E<W-$=&P@/2!0<F]C97-S1'1L#0H)"0D)
M"2,@=V4@;F]W(&AA=F4@=&AE(&)A<V4@:6YF;W)M871I;VX@=&AA="!W92!N
M965D('1O(&9U;&9I;&P-"@D)"0D)(R!A(')E<75E<W0 at 9F]R('!R;V-M86X@
M:6YF;W)M871I;VXN#0H)"0D)96QS93H-"@D)"0D)(R!5;F1E=&5R;6EN960@
M;65S<V%G92!T>7!E(&9R;VT@<')O8VUA;@T*"0D)"0DC(&YE960@=&\@<F5P
M;W)T(&%N(&5R<F]R+ at T*"0D)"0EP<FEN=" B56YA8FQE('1O(&EN:71I86QI
M>F4@<')O8VUA;B Q(@T*"0D)"0T*"0EX/3 -"@D)=VAI;&4@,3H-"@D)"6-H
M96-K7W)T<FX@/2!0>5!M+F-H96-K<2A0<F]C;6%N44ED+$9R9611260L<V5L
M9BYP<F5A9"Q04D]#34%.7U1)345/550I#0H)"0EX/7 at K,0T*"0D)<')I;G0@
M(DEN($QO;W B("P@> T*"0D):68@='EP92AC:&5C:U]R=')N*2 ]/2!4=7!L
M951Y<&4Z#0H)"0D)(R!N965D('1O(&-H96-K(%LP72!F;W(@;65S<V%G92!T
M>7!E#0H)"0D);65S<V%G95]T>7!E(#T at 8VAE8VM?<G1R;ELP70T*"0D)"6EF
M(&UE<W-A9V5?='EP92 ]/2 P. at T*"0D)"0E0>5!M+G!R;V-H9'(H<V5L9BY0
M<F]C;6%N44ED+'-E;&8N1G)E9%%)9"D-"@D)"0D)(R!W92!H860 at 82!T:6UE
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M<V5L9BY0<F]C97-S2&1R. at T*"0D)"0D);7-G;&ES=#TH8%!23T-315)68"Q@
M1E)%1%-%4E9 at +"=9)RQ@;65S<V%G95]T>7!E8"LG82<I#0H)"0D)"0EM<V<]
M:F]I;BAM<V=L:7-T+"(K*RLB*0T*"0D)"0D);7-G;&ES=#UM<V<L(GQ\(@T*
M"0D)"0D);7-G/7-T<FEN9RYL:G5S="AJ;VEN*&US9VQI<W0L(B(I+$U31U],
M14Y'5$@I#0H)"0D)"0ES96QF+G%U975E+G!U="A0<F]C97-S2&5A9&5R*0T*
M"0D)"0D);G5M7W=R:71T96X];W,N=W)I=&4H<V5L9BYP=W)I=&4L;7-G*0T*
M"0D)"0D)<V5L9BY0<F]C97-S2&1R(#T at 4')O8V5S<TAE861E<@T*"0D)"0EI
M9B!0<F]C97-S1'1L("$]('-E;&8N4')O8V5S<T1T;#H-"@D)"0D)"2,@=V4@
M;F5E9"!T;R!S96YD(&$@;65S<V%G92!T;R!F<F5D#0H)"0D)"0EM<V=L:7-T
M/2A at 4%)/0U-%4E9 at +&!&4D5$4T525F L)UDG+&!M97-S86=E7W1Y<&5@*R=B
M)RD-"@D)"0D)"6US9SUJ;VEN*&US9VQI<W0L(BLK*R(I#0H)"0D)"0EM<V=L
M:7-T/6US9RPB?'PB#0H)"0D)"0EM<V<]<W1R:6YG+FQJ=7-T*&IO:6XH;7-G
M;&ES="PB(BDL35-'7TQ%3D=42"D-"@D)"0D)"7-E;&8N<75E=64N<'5T*%!R
M;V-E<W-(96%D97(I#0H)"0D)"0EN=6U?=W)I='1E;CUO<RYW<FET92AS96QF
M+G!W<FET92QM<V<I#0H)"0D)"0ES96QF+E!R;V-E<W-$=&P@/2!0<F]C97-S
M1'1L#0H)"0D)96QI9B!M97-S86=E7W1Y<&4@/3T@,CH-"@D)"0D)(R!W92!H
M879E(&$@<')O8V5S<R!D971A:6P@<F5Q=65S=" -"@D)"0D)4')O8V5S<TEN
M9F\@/2!C:&5C:U]R=')N6S%=#0H)"0D)"51O475E(#T at 1'1L4F5Q1&EC=%M0
M<F]C97-S26YF;ULP75T-"@D)"0D);7-G;&ES=#TH8%!23T-315)68"Q at 1E)%
M1%-%4E9 at +"=9)RQ@;65S<V%G95]T>7!E8"LG(2<K5&]1=64I#0H)"0D)"6US
M9SUJ;VEN*&US9VQI<W0L(BLK*R(I#0H)"0D)"6US9VQI<W0];7-G+")\?"(-
M"@D)"0D);7-G/7-T<FEN9RYL:G5S="AJ;VEN*&US9VQI<W0L(B(I+$U31U],
M14Y'5$@I#0H)"0D)"7-E;&8N<75E=64N<'5T*%!R;V-E<W-);F9O*0T*"0D)
M"0EN=6U?=W)I='1E;CUO<RYW<FET92AS96QF+G!W<FET92QM<V<I#0H-"@D)
M"0EE;&EF(&UE<W-A9V5?='EP92 ]/2 S. at T*"0D)"0E.;VYE#0H)"0D)"2,@
M=V4@:&%D(&$@8V]N9FEG(')E9G)E<V@@97)R;W(@#0H)"0D)96QI9B!M97-S
M86=E7W1Y<&4@/3T at -#H-"@D)"0D)3F]N90T*"0D)"0DC('=E(&AA9"!A(')E
M<W1A<G0@<F5Q=65S="!E<G)O<@T*"0D)"65L:68@;65S<V%G95]T>7!E(#T]
M(#4Z#0H)"0D)"4YO;F4-"@D)"0D)(R!W92!H860 at 82!S:'5T9&]W;B!R97%U
M97-T(&5R<F]R( T*"0D)"65L:68@;65S<V%G95]T>7!E(#T](#DZ#0H)"0D)
M"2,@=V4@;F5E9"!T;R!R96%D(&]U<B!P<F]C97-S('!I<&4-"@D)"0D)9F0]
M<V5L9BYP<F5A9 T*"0D)"0ED871A/6]S+G)E860H9F0L35-'7TQ%3D=42"D-
M"@D)"0D);65S<V%G97,]<W!L:70H9&%T82PG?'PG*0T*"0D)"0EF;W(@;7-G
M(&EN(&UE<W-A9V5S. at T*"0D)"0D)9G0])R<-"@D)"0D)"71T/2<G#0H)"0D)
M"0EQ:3TG)PT*"0D)"0D)<&%Y;&]A9#TG)PT*"0D)"0D):68@;7-G. at T*"0D)
M"0D)"69T+'1T+'%I+'!A>6QO860]<W!L:70H;7-G+"(K*RLB*0T*"0D)"0D)
M:68 at 9G0@/3T at 8$92141#3$E%3E1 at .@T*"0D)"0D)"2,@=V4 at 87)E(&=E='1I
M;F<@82!R97%U97-T(&9O<B!A9&1I=&EO;F%L(&EN9F\-"@D)"0D)"0ER97%?
M='EP92 ]('!A>6QO861;,%T-"@D)"0D)"0ED971A:6P@/2!P87EL;V%D6S$Z
M70T*"0D)"0D)"6EF(')E<5]T>7!E(#T](#(Z#0H)"0D)"0D)"2,@=V4@:&%V
M92!A(')E<75E<W0 at 9F]R(&-O;F9I9R!R969R97-H#0H)"0D)"0D)"5!Y4&TN
M<')O8V1T;"AS96QF+E!R;V-M86Y1260L<V5L9BY&<F5D44ED+ T*"0D)"0D)
M"0D)"7-E;&8N<')E860L9&5T86EL*0T*"0D)"0D)"65L:68@<F5Q7W1Y<&4@
M/3T@,SH-"@D)"0D)"0D)(R!W92!H879E(&$@<F5Q=65S="!F;W(@8V]N9FEG
M(')E9G)E<V at -"@D)"0D)"0D)4$YU;2Q4;U%U92 ]('-P;&ET*&1E=&%I;"PG
M*R<I#0H)"0D)"0D)"7-E;&8N1'1L4F5Q1&EC=%M03G5M72 ](%1O475E#0H)
M"0D)"0D)"5!Y4&TN8V]N9FEG*'-E;&8N4')O8VUA;E%)9"QS96QF+D9R9611
M260L#0H)"0D)"0D)"0D)<V5L9BYP<F5A9"Q03G5M*0T*"0D)"0D)"65L:68@
M<F5Q7W1Y<&4@/3T at -#H-"@D)"0D)"0D)(R!W92!H879E(&$@<F5Q=65S="!F
M;W(@<F5S=&%R= T*"0D)"0D)"0E03G5M+$AA<F13;V9T/7-P;&ET*&1E=&%I
M;"PB*R(I#0H)"0D)"0D)"5!Y4&TN<F5S=&%R="AS96QF+E!R;V-M86Y1260L
M<V5L9BY&<F5D44ED+ T*"0D)"0D)"0D)"7-E;&8N<')E860L4$YU;2Q(87)D
M4V]F="D-"@D)"0D)"0EE;&EF(')E<5]T>7!E(#T](#4Z#0H)"0D)"0D)"2,@
M=V4@:&%V92!A(')E<75E<W0 at 9F]R('-H=71D;W=N( T*"0D)"0D)"0DC('1H
M92!D971A:6P@:&5R92!I<R!(87)D(&]R(%-O9G0-"@D)"0D)"0D)4'E0;2YS
M:'5T9&]W;BAS96QF+E!R;V-M86Y1260L<V5L9BY&<F5D44ED+ T*"0D)"0D)
M"0D)"7-E;&8N<')E860L9&5T86EL*0T*#0H)"0D)96QS93H-"@D)"0D)(R!5
M;F1E=&5R;6EN960@;65S<V%G92!T>7!E(&9R;VT@<')O8VUA;@T*"0D)"0DC
M(&YE960@=&\@<F5P;W)T(&%N(&5R<F]R+ at T*"0D)"0EP<FEN=" B56YA8FQE
M('1O(&EN:71I86QI>F4@<')O8VUA;B R(@T*"0D)"0T*"61E9B!E>&ET*'-E
M;&8I. at T*"0EP<FEN="@G26X at 17AI="<I#0H)"6EF('-E;&8N1G)E9%%)9#H-
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M=VEL;"!B92!S:&%R960 at 8GD@86QL(&]F('1H92!P<F]C97-S('1H<F5A9',@
M8F5I;F<@:6YS=&%T:6%T960-"B,@(R,C(R,C(R,C(R,C(R,C(R,C(R,C(R,C
M(R,C(R,C(R,C(R,C(R,C(R,C(R,C(R,C(R,C(R,C(R,C(R,C(R,C(R,C(R,C
M(R,C(R,C(R C#0H-"B, at 3&]C86P@=F%R:6%B;&4@:6YI=&EA;&EZ871I;VX-
M"@EF:7)S=%]T:6UE(#T@, T*#0H)<F5A9%]D:6-T(#T@>WT-"@EW<FET95]D
M:6-T(#T@>WT-"@E1=65U95]D:6-T(#T@>WT-"@T*"6QO8V%L9FEL96QI<W0@
M/2!;70T*"71H<F5A9&QI<W0@/2!;70T*#0H);%%U975E(#T at 475E=64H,3 I
M#0H)9E%U975E(#T at 475E=64H,3 I#0H)85%U975E(#T at 475E=64H,3 I#0H)
M;VU1=65U92 ](%%U975E*#$P*0T*"6AA<F11=65U92 ](%%U975E*#$P*0T*
M"65X:691=65U92 ](%%U975E*#$P*0T*"7-C475E=64@/2!1=65U92 at Q,"D-
M"@EC:&%N475E=64@/2!1=65U92 at Q,"D-"@EP<F]C475E=64@/2!1=65U92 at Q
M,"D-"@E1=65U95]D:6-T6TQ)4U1%3E-%4E9=(#T@;%%U975E#0H)475E=65?
M9&EC=%M&4D5$4T525ET@(" ](&91=65U90T*"5%U975E7V1I8W1;04Q!4DU3
M15)672 @/2!A475E=64-"@E1=65U95]D:6-T6T]-4T525ET@(" @(#T@;VU1
M=65U90T*"5%U975E7V1I8W1;3U-315)672 @(#T@:&%R9%%U975E#0H)475E
M=65?9&EC=%M%6$E&4T525ET@(" ](&5X:691=65U90T*"5%U975E7V1I8W1;
M4T-315)672 @(" @/2!S8U%U975E#0H)475E=65?9&EC=%M#2$%.4T525ET@
M(" @(#T at 8VAA;E%U975E#0H)475E=65?9&EC=%M04D]#4T525ET@(" @(#T@
M<')O8U%U975E#0H-"@DC('-T87)T(&)E('-E=&EN9R!U<"!A('-O8VME="!F
M;W(@;&ES=&5N#0H)(W,@/2!S;V-K970N<V]C:V5T*$%&7TE.150L(%-/0TM?
M4U1214%-*0T*"2-S+F)I;F0H2$]35"Q30TE07U!/4E0I#0H)(W,N;&ES=&5N
M*#$I#0H)<SU.;VYE#0H-"@T*"2-#<F5A=&4@=&AE(%!R;V-M86X at 4V5R=F5R
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M+%M=+&QO8V%L9FEL96QI<W0L,2D-"@D):68 at 83H-"@D)"69O<B!T:')E860@
M:6X@=&AR96%D;&ES=#H-"@D)"0ET:')E860N97AI="@I#0H)"0EE>&ET*# I
C#0H)"7!R:6YT(")O=71E<B!,;V]P(BQX#0H)"7@]>"LQ#0H`
`
end




From aa8vb at yahoo.com  Fri Nov 19 13:44:22 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Fri, 19 Nov 1999 13:44:22 -0500
Subject: Building wxPython [was Re: wxPython and DISLIN work nicely together]
In-Reply-To: <002301bf32b5$97a5d440$0301a8c0@easystreet.com>
References: <80s0b4$g5n@mail.psy.uva.nl> <19991117090214.B807152@vislab.epa.gov> <017a01bf31f8$6bbea590$1a25d2d1@jenkondev.com> <19991119064944.A996886@vislab.epa.gov> <002301bf32b5$97a5d440$0301a8c0@easystreet.com>
Message-ID: <19991119134422.A1004588@vislab.epa.gov>

Robin Dunn:
 |>  |> If only wxPython would build on IRIX...
 |>
 |What is the output of "wx-config --cflags" ?  It should have the right
 |include directories for getting all the wxWindows headers.
 |
 |Oh wait a minute, it says "/bin/sh: wx-config:  not found" so it looks like
 |you didn't do a "make install" for wxGTK.

To the latter, I've already done "make install" for wxGTK:

     > ls ~/software/wxGTK-2.1.11/
     bin/      include/  lib/      share/

and wx-config does exist, in the $WXWIN directory:

    > python $WXWIN/utils/wxPython-2.1.11/distrib/build.py -b MAKE=gmake
    ...
    /bin/sh: wx-config:  not found
    ...
    > cd ../../..
    > find . -name wx-config
    ./wx-config

And to the former, here's my 'wx-config --cflags' output:
 
    > wx-config --cflags
    -I/home/rhh/software/wxGTK-2.1.11/include -I/home/rhh/software/wxGTK-2.1.11/lib/wx/include -D__WXMOTIF__ -I/usr/Motif-1.2/include


Leafing through build.py, the problem is that wxPython assumes wxGTK's
wx-config is in $PATH.  I don't have root so I don't install things in
/usr/local/*.  (I also don't like having package intermingle in a shared
directory tree; too hard to remove them later.)  So the change to hack
around this is to temporarily add $WXWIN's bin directory to $PATH:

     env PATH=$WXWIN/bin:$PATH /usr/local/bin/python \
         $WXWIN/utils/wxPython-2.1.11/distrib/build.py -b MAKE=gmake

Perhaps wxPython should have a --with-wxwindows= option, or take it's
queues off of $WXWIN?

It then went on a little further to try and compile helpers.cpp, but failed
with 84 errors, all (or almost all) due to a bad NULL comparisons in the
wxWindows header files.  NULL is defined as ((void*)0) and cannot be
assigned to / compared with another type without a cast in C++.  As this is
C++, I believe they should be using "0" for assignments and comparisons.

Odd that they even note this fact in wxWindows' defs.h file:

   #if defined(__SGI_CC__)
   // Okay this is bad styling, but the native SGI compiler is very picky, it
   // wont let you compare/assign between a NULL (void *) and another pointer
   // type. To be really clean we'd need to pass in another argument, the type
   // of p.

-- 
Randall Hopper
aa8vb at yahoo.com



From steveef at att.net  Tue Nov 30 01:37:47 1999
From: steveef at att.net (steveef at att.net)
Date: 30 Nov 1999 06:37:47 GMT
Subject: Get Paid While You Surf The Web!!!  9438
Message-ID: <81vrbr$bt3$7247@bgtnsc03.worldnet.att.net>

Get paid while you surf the web here: http://www.alladvantage.com/home.asp?refid=FCJ475


cinb




From tlane at parlant.com  Tue Nov  9 10:50:47 1999
From: tlane at parlant.com (Thomas Lane)
Date: Tue, 09 Nov 1999 08:50:47 -0700
Subject: Image buttons with Tkinter
References: <38270596.A391995C@parlant.com> <01fd01bf2a3b$52cbd3b0$f29b12c2@secret.pythonware.com>
Message-ID: <382842D7.91D729F2@parlant.com>

Thanks for the help Fredrik.
I am new to Python, so this particular effect of scoping was not
apparent to me at first.
Since I was doing this inside a method I changed my code to look like:

# Put remove.gif on a button (actionbar is a frame for various controls)
self.image_delete = PhotoImage(file="remove.gif")
deletebutton = Button(actionbar,image=image_delete)
deletebutton.place(x=0)

That solved the problem.
-- 
Thomas Lane
Provo, UT
tlane at parlant.com

Fredrik Lundh wrote:
> 
> Thomas Lane wrote:
> > I'm trying to create a button that displays an image instead of text,
> > but I don't seem to be having any luck. The button just comes up gray.
> 
> see:
> 
>     http://www.python.org/doc/FAQ.html#4.69
>     4.69. Images on Tk-Buttons don't work in Py15?
> 
> </F>
> 
> <!-- (the eff-bot guide to) the standard python library:
> http://www.pythonware.com/people/fredrik/librarybook.htm
> -->



From tismer at appliedbiometrics.com  Tue Nov 16 12:14:44 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Tue, 16 Nov 1999 18:14:44 +0100
Subject: list.without()?
References: <NDBBIAPHEKGOBIHCBLIICEOMCAAA.mcfletch@vrtelecom.com> <y0jiu32iyi8.fsf@vier.idi.ntnu.no>
Message-ID: <38319104.F54AECEF@appliedbiometrics.com>


"Magnus L. Hetland" wrote:
> 
> "Mike Fletcher" <mcfletch at vrtelecom.com> writes:
> 
> > Well...
> >
> > def without( source, element):
> >       temp = source[:]
> >       while temp:
> >               try:
> >                       temp.remove( element )
> >               except:
> >                       break
> >       return temp
> 
> Whoa!
> 
> Why on *earth* would you use a while/try/break-combination? IMO a much
> more natural solution would be:
> 
> def without(source, element):
>     result = source[:]
>     if element in result:
>         result.remove(element)
>     return result

But that would be wrong.
"Without" means "there is no such object in it", while list.remove
removes the first occourance of an object.

Mike's implementation appears to be very cheap, since through
the try..except clause, he can combine removal with test
for existance.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From skip at mojam.com  Tue Nov  2 14:19:02 1999
From: skip at mojam.com (Skip Montanaro)
Date: Tue, 2 Nov 1999 13:19:02 -0600 (CST)
Subject: Assorted Questions regarding the future of Python...
In-Reply-To: <oqogddnep5.fsf@titan.progiciels-bpi.ca>
References: <6D8A17398E28D3119F860090274DD7DB49833A@pces.cadlab.it>
	<006201bf2517$50edf1f0$f29b12c2@secret.pythonware.com>
	<oqogddnep5.fsf@titan.progiciels-bpi.ca>
Message-ID: <14367.14630.983098.834929@dolphin.mojam.com>

>>>>> "Francois" == ISO-8859-1  <ISO-8859-1> writes:

    Fran?ois> I'm curious.  Does Python use standard optimisation techniques
    Fran?ois> on that byte code (CSE, strength reduction, and all that)?  I
    Fran?ois> would be tempted to guess that the stack machine receives code
    Fran?ois> that pretty closely resemble the source code.  Would the stack
    Fran?ois> machine be able to express optimisations?  For example, could
    Fran?ois> we copy or exchange arbitrary stack elements?

Several people have worked on various optimizations to the Python virtual
machine or to the generated byte code over time.  Some of it has already
been adopted into compile.c or ceval.c.  As a result, I believe the latest
version of Python is probably at least twice as fast as 1.3 ('twould be
kinda neat if someone did the work to benchmark pystone against various
versions of the interpreter).  For pointers to stuff I've done (including a
peephole optimizer for Python bytecode), check out

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

For other reported work, check out the Python Conferece proceedings that are
online.  Pointers are available at

    http://www.python.org/workshops/

For a completely different look at the Python virtual machine, check out
the Rattlesnake list:

    http://www.egroups.com/group/rattlesnake/

Rattlesnake hasn't gone anywhere in about a year, due to lack of time on my
part.  The idea is to implement a new three address virtual machine using
the existing runtime stack as the "registers", then convert the existing
stack-based bytecode to three address bytecode.  Three address form is
easier to optimize and should (in theory) execute faster after you apply
optimizations.  If people are interested in fiddling with it I'd be happy to
generate a context diff against the current CVS repository and put it up on
my Python Bits page.

Skip Montanaro | http://www.mojam.com/
skip at mojam.com | http://www.musi-cal.com/
847-971-7098   | Python: Programming the way Guido indented...



From nascheme at enme.ucalgary.ca  Tue Nov 30 00:23:17 1999
From: nascheme at enme.ucalgary.ca (Neil Schemenauer)
Date: Tue, 30 Nov 1999 05:23:17 GMT
Subject: performance of tuple-less Python?
References: <1268226961-30285235@hypernet.com>
Message-ID: <slrn846nr4.kpg.nascheme@adler.dynodns.net>

Gordon McMillan <gmcm at hypernet.com> wrote:
>They're only redundant if the concept of mutability is 
>meaningless. A reference-based language without a concept 
>of mutability would be, um, an adventure.

Indeed.

>what-if-six-turned-out-to-be-nine-ly y'rs

Wasn't there a version of Fortran that allowed this?  Python is
not totally with a means of enjoying this adventure.

    >>> None = 1
    >>> a = None
    >>> if a:     
    ...     print "oh no"
    ... 
    oh no

That could be fun (or dangerous) in a Python project with lots of
developers.


    Neil (professional hacker at work, do not attempt at home :)



From nospam at bitbucket.com  Sun Nov 28 03:06:49 1999
From: nospam at bitbucket.com (Phil Mayes)
Date: Sun, 28 Nov 1999 00:06:49 -0800
Subject: is there a safe/easy way to search and replace text w/in an html page?
References: <3840593B.2D1A2E30@earthlink.net>
Message-ID: <Ho504.65$E3.5775@newsfeed.avtel.net>

Levi Cook wrote in message <3840593B.2D1A2E30 at earthlink.net>...
>Hello,
>
>I'm new to python and I've been tasked w/ writing a script that fetches
>a web-page and highlights the term that the user searched for. Passing
>the URL and search term to my script was a cinch and fetching the page
>was also quite easy.
>
>The crux of my problem lies with establishing a strategy for searching
>out the users term and replacing it with css tags surrounding the
>original term.
>At first glance the re module seemed sufficient (and maybe still is ),
>but it doesn't really have any knowledge of HTML semantics.

Wrapping the found text in <TAG of your choice> should work fine.
One caveat is if tags appear within the target text, eg Netscape pages
used to bump the font size of the first letter of a title by one:
  <FONT SIZE=+1>N</FONT>etscape

>My next
>venture was to look at the htmllib module. This appears to parse the
>html fine, but i'm still at a bit of a loss on what this actually does
>for me, particularly how do I get my parser to output my page.
>
>Does anyone know where I can find code that does anything remotely like
>this?
>Does anyone know where I can find good reference on the htmllib module,
>the python library reference is pretty bleak in this area.

> [code sample snipped]

The htmllib module is fairly primitive, eg it doesn't do tables, frames,
images or css, so unless you are able to control the HTML contents
pretty strictly, you would be better off asking a competent browser to
display your modified page - write it somewhere and load file://...

What browser to talk to, and how, is going to be machine-specific.
On Win98, Internet Explorer can be controlled with PythonWin.
Look at win32com\test\testExplorer.py
--
Phil Mayes    pmayes AT olivebr DOT com
Olive Branch Software






From jima at aspectdv.com  Tue Nov  9 22:04:15 1999
From: jima at aspectdv.com (Jim Althoff)
Date: Tue, 09 Nov 1999 19:04:15 -0800
Subject: __str__ vs. __repr__
In-Reply-To: <slrn82hhtv.ppi.neelk@brick.cswv.com>
References: <000401bf25b9$88ea57e0$2e2d153f@tim>
 <199911040157.UAA11908@eric.cnri.reston.va.us>
 <046101bf26d4$2dd1a720$01ffffc0@worldnet.att.net>
Message-ID: <4.2.0.58.19991109190103.00afcab0@mail.aspectdv.com>

I agree with Neel.  The interactive interpreter is an incredibly
powerful tool for learning Python and for developing code.
And with JPython, it is a wonderful, powerful tool
for learning the Java APIs as well.

Jim



At 01:17 AM 11/10/99 +0000, Neel Krishnaswami wrote:
>Emile van Sebille <emile at fenx.com> wrote:
> >
> >It seems to me that issues arise when using the interactive interpreter.
> >I suspect that little useful work is done in this mode, and is used
> >predominately for quick testing while creating python programs.
>
>I'd like to politely disagree . Combined with Tim Peters and Barry
>Warsaw's emacs py-mode, Python's interactive mode is the single most
>useful feature of the language for me. With it, I can test each and
>every method, loop, and statement as I code. This does wonders for my
>understanding of my code. And not incidentally, reliability goes way
>up as well.
>
>I learned Perl before Python, but my Python skill surpassed my Perl
>skill after a much smaller time investment. I lay this pretty much
>wholely at the feet of the interactive interpreter -- I could find out
>the answers to puzzles without a tedious edit-save-run cycle.
>
>
>Neel
>
>--
>http://www.python.org/mailman/listinfo/python-list





From akuchlin at mems-exchange.org  Thu Nov  4 17:06:12 1999
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: 04 Nov 1999 17:06:12 -0500
Subject: Python Conference Hotel Change
References: <199911032224.RAA10668@eric.cnri.reston.va.us> <3824f510.12238077@news.demon.co.uk>
Message-ID: <3dn1stvrmz.fsf@amarok.cnri.reston.va.us>

andy at robanal.demon.co.uk (Andy Robinson) writes:
> Guido van Rossum <guido at cnri.reston.va.us> wrote:
> >desirable facility for the conference.  The new facility is the Key
> >Bridge Marriott in Arlington, Virginia.  The Key Bridge Marriott is
> >conveniently located on the Virginia side of the Key Bridge in the
>
> My memory may be fuzzy, but isn't it an even shorter stroll from there
> to Bardo, one of the world's biggest microbreweries?

It's quite a lengthy stroll, really, but a 5- or 10-minute car ride.
Doubtless an evening trip can be arranged; I just hope it'll involve
playing pool.  (I'd be astounded to learn that Bardo's is the largest
*anything*, though; it seems to be there are other, larger breweries
in the area.)

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
Well, that's a little thing -- the specification.
    -- Guido van Rossum at IPC7




From pinard at iro.umontreal.ca  Mon Nov  1 13:45:21 1999
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 01 Nov 1999 13:45:21 -0500
Subject: Detect file existance before open?
In-Reply-To: "Gordon McMillan"'s message of "Sat, 16 Oct 1999 23:21:43 -0400"
References: <1271992746-7662859@hypernet.com>
Message-ID: <oqzowygif2.fsf@titan.progiciels-bpi.ca>

"Gordon McMillan" <gmcm at hypernet.com> writes:

> > I would like to check for the existance of a file before I
> > attempt an open and maybe throw an exception.

> os.path.exists(nm), probably also os.path.isfile(nm).

And also os.path.islink(nm), unfortunately.

If a symbolic link exists and points to a deleted file, say, os.path.exists
returns None.  I found out that you need to test both os.path.exists and
os.path.islink before assuming the path is free for safe re-use.

It is a bit of constant annoyance in some projects, and do not clearly see
how Python `os' library could be improved to make this more transparent.
I guess we just have to learn to live with this (?) ...

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





From mwh21 at cam.ac.uk  Wed Nov 10 03:46:19 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 10 Nov 1999 08:46:19 +0000
Subject: MSWIN:getch() UNIX:?
References: <80a3rk$dg4$1@nslave2.tin.it>
Message-ID: <m3vh7an38k.fsf@atrus.jesus.cam.ac.uk>

"Francesco" <frcontin at tin.it> writes:

> There's a module with a getch() function that can run on all
> platforms?

This is a regular. Have you tried dejanews?

In summary - there's a version that'll work on almost every platform
your likely to come across, but isn't guaranteed to (it's unlikely to
work if the terminal actually is a teletype jobby printing on paper,
I'd imagine).

Happy searching!
Michael



From greg.ewing at compaq.com  Wed Nov  3 09:26:28 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Thu, 04 Nov 1999 03:26:28 +1300
Subject: A Comparison of Python and Ruby Extension Interfaces (Re: Comparison 
 between Python and "Ruby")
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp> <7vjtrc$hmm$1@newshost.accu.uu.nl>
Message-ID: <38204614.2EAC6051@compaq.com>

Martijn Faassen wrote:
> 
> From what I hear in this thread its C interfacing would also be interesting
> to compare with Python's, especially since Ruby uses garbage collection,
> which is something Python's looking at using in the future, in some
> way.

Okay, here are some thoughts following a brief time looking
at and comparing the "dbm" module implementations from Ruby
and Python.

The first observation is that the Ruby version is 13K in
size, whereas the Python version is only about 7.5K.
However, this is not a fair comparison, because the Ruby
version defines more methods.

Eyeballing the code, I must say that the Ruby version
generally looks neater and tidier. Some reasons for this
seem to be:

* The names of the Ruby interface routines are often
shorter than the equivalent Python ones.

* Checking types of arguments and raising errors seems to
be easier in the Ruby version, e.g. Ruby

    Check_Type(keystr, T_STRING);
    key.dptr = RSTRING(keystr)->ptr;
    key.dsize = RSTRING(keystr)->len;

vs. Python

    if ( !PyArg_Parse(v, "s#", &krec.dptr, &krec.dsize) ) {
        PyErr_SetString(PyExc_TypeError,
            "dbm mappings have string indices only");
        return -1;
    }

The Ruby version isn't all that much shorter, but it
looks neater. On the other hand, the Python version manages
to produce a more specific error message; if the Ruby
version did that too, it would be longer.

* Since Ruby has mark-and-sweep garbage collection, there
is no manipulation of reference counts. There weren't
actually any examples of this in the Python dbm module,
though.

* Ruby seems to be able to raise exceptions without
having to explicitly propagate them back out of routines,
which eliminates a lot of result testing and returning of
-1's and None's that goes on in Python extension code.

Some other observations, in no particular order:

* In Ruby, routines with a fixed number of arguments can
be defined so that they are received as separate C
arguments instead of a vector, e.g.

static VALUE
fdbm_store(obj, keystr, valstr)
    VALUE obj, keystr, valstr;

This doesn't actually save much code, however, since you
still have to test and convert each argument.

* Ruby's method of representing an extension type is a
whisker more complicated. Instead of allocating a single
object, you have to create an object of type DATA, which
contains a pointer to a class and a pointer to a C
structure.

* The mechanism for associating methods with object types
is somewhat simpler in Ruby. In Python, some methods live
in slots in the type object, while others live in a method
table or are somehow otherwise accessed through the getattr
method. In Ruby, all methods, including the ones for built
in operations, are attached to the object's class, which seems
to play the role of both the Python type object and method
table. The initialisation function of the Ruby dbm module
starts off like this:

void
Init_dbm()
{
    cDBM = rb_define_class("DBM", rb_cObject);
    rb_include_module(cDBM, rb_mEnumerable);

    rb_define_singleton_method(cDBM, "open", fdbm_s_open, -1);
    rb_define_singleton_method(cDBM, "new", fdbm_s_open, -1);
    rb_define_method(cDBM, "close", fdbm_close, 0);
    rb_define_method(cDBM, "[]", fdbm_fetch, 1);
    rb_define_method(cDBM, "[]=", fdbm_store, 2);
    rb_define_method(cDBM, "indexes",  fdbm_indexes, -1);

etc. There is no need to declare and initialise a type object
or method table as in the Python version, which makes things
somewhat more direct and easier to follow.

Summary:

Ruby extension modules seem to be slightly easier to write and
read than Python ones. The main reasons for this are the lack
of reference counting, a less obtrusive exception propagation
mechanism, and a more uniform notion of types and classes.
There are also some minor conveniences, such as C methods which
take an ordinary argument list instead of a vector. The
broad approach to exension writing is, however, fairly similar
in both systems.

Tomorrow I may take a deeper look at how Ruby unifies types
and classes and see whether there is anything there that
Python could learn from.

Greg



From wilsher at my-deja.com  Fri Nov 12 12:18:36 1999
From: wilsher at my-deja.com (wilsher at my-deja.com)
Date: Fri, 12 Nov 1999 17:18:36 GMT
Subject: UML tools: Tcl or Python?
References: <382A3896.7FD4F3B0@objectsbydesign.com>
Message-ID: <80hi56$97$1@nnrp1.deja.com>

In article <382A3896.7FD4F3B0 at objectsbydesign.com>,
  Stuart Zakon <info at objectsbydesign.com> wrote:
> Interesting developments for scripting in UML tools:
>
> First ObjectDomain switched its scripting language from Tcl to
JPython.
> Next Together/J 3.0 switched its scripting language from JPython to
Tcl.
>

Hmmm, are you sure about this? In the 3.0 Beta they were still
supporting
JPython, but de-emphasizing it in favour of the Java API. I just had a
look at the 3.0 final help docs, which states that "a script is a Java
class". There is no mention of either JPython or TCL, as far as I can
see.
Needless to say, I think JPython makes for a much better application
scripting language than Java....

--Thomas


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



From vroonhof at frege.math.ethz.ch  Sun Nov 28 09:37:04 1999
From: vroonhof at frege.math.ethz.ch (Jan Vroonhof)
Date: 28 Nov 1999 15:37:04 +0100
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <867lj85g4c.fsf@g.local> <al8yabo9fm3.fsf@sirppi.helsinki.fi> <86so1v4jlj.fsf@g.local> <al8hfiaan8p.fsf@sirppi.helsinki.fi> <86ln7m2t31.fsf@g.local> <m3emddrmuh.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <byln7iof73.fsf@urysohn.math.ethz.ch>

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

> This leads to a thought: would it be possible using lisp's silly
> amount of flexibility to get the reader to auto-paranthesize (word?
> doubt it):

It must be in the air or something ;-) I had the same though a while a
ago. I even have got a name LISI (Lot's of Significant Indentation)

The rules could be extremely simple.
1. each new indentation level
introduces another pair of parenthesis
2. Each new line is wrapped in parenthesis

Then you get things like

defun count-words-region (start end &optional buffer)
  interactive "_r"
  save-excursion
    set-buffer (or buffer (current-buffer))
    let
        words 0
      goto-char start
      while (< (point) end)
        when (forward-word 1)
          incf words
      when  (interactive-p) 
        message "Region has %d words" words
      identity words

That last line is ugly because in this way it is very difficult to
distinguish between a no-argument function call and a variable
reference.

Sometimes however I think it would be more fashionable to replace all
parenthesis by angle brackets and re-brand Lisp as LIML.

Jan






From fdrake at acm.org  Tue Nov 16 15:28:27 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Tue, 16 Nov 1999 15:28:27 -0500 (EST)
Subject: Final classes like in java???
In-Reply-To: <14385.48236.202756.297241@goon.cnri.reston.va.us>
References: <80sao7$1848$1@rtpnews.raleigh.ibm.com>
	<14385.45862.55010.146256@weyr.cnri.reston.va.us>
	<14385.48236.202756.297241@goon.cnri.reston.va.us>
Message-ID: <14385.48747.331513.47679@weyr.cnri.reston.va.us>

I wrote:
 > Not guaranteed, but this should be good enough:

Jeremy Hylton writes:
 > This doesn't really work.

  That's exactly why I said it wasn't guaranteed.  ;-)


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From thomas at bibsyst.no  Mon Nov  1 06:25:55 1999
From: thomas at bibsyst.no (Thomas Weholt)
Date: Mon, 01 Nov 1999 12:25:55 +0100
Subject: Just a note
References: <381D5E84.98931682@bibsyst.no>
Message-ID: <381D78C3.9BBECCB0@bibsyst.no>

Hi,

I could use stuff like MySQL too, if it`s easy to install and use with
Python, but since I only want to store objects and fetch them by a single
key, the simplest way to do this would be using some of the built-in stuff
in Python.

I will store HUGE amounts of data, so I think Gadfly would be "too small"
as DBMS, since it stores all it`s data in memory ( as I recall ) .

PS! If this mail made no sense, try reading the message prior to this one.

Best regards,
Thomas Weholt




From ivanlan at callware.com  Tue Nov  9 00:22:55 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Mon, 08 Nov 1999 22:22:55 -0700
Subject: python: something completely different
References: <3820C0B8.62B2@iowegian.com> <38249ED2.BB7EA1DE@znet.com> <8028jv$7cf$1@hub.org> <38259F18.437@iowegian.com> <38278B5F.1AFF953C@python.net> <mENV3.6280$ra7.129744@news1.wwck1.ri.home.com>
Message-ID: <3827AFAF.FF46C33@callware.com>

Hi All--

bob yazon wrote:
> 
> The orig post got me interested in the stuff. I just finished installing it
> + win32ext, numeric, and gnuplot.  I can get pretty good graphs using
> gnuplot, I think I'll start using python instead of Mlab.
> 
> --ry
> 

<g>  Would you consider letting us in on your secrets?  You could write
it up and it could go in the FAQ.

<never-announce-your-secret-vices-to-the-world-unless-you-want-publicity>-ly
y'rs,
Ivan;-)
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours
----------------------------------------------



From mikael at isy.liu.se  Fri Nov  5 04:06:17 1999
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Fri, 05 Nov 1999 11:06:17 +0200 (MET)
Subject: Apply a function to each list member?
In-Reply-To: <slrn824gee.1f7.bwinton@tor.dhs.org>
Message-ID: <XFMail.991105110617.mikael@isy.liu.se>

On 05-Nov-99 Blake Winton wrote:
 >  Speaking as someone who's one language is English, "Probably not".
 >  "map" is really a Lisp/Scheme concept, and bears only a passing
 >  relationship to English.  (Huh, what do ya know, looking up "map"
 >  on http://www.m-w.com/ reveals (in meaning 4) that "map" actually
 >  means "FUNCTION".  I have no idea why they capitalized it.)
 
I guess the terminology originally is taken from mathematics. A map is 
a mathematical concept, similar to a function. Perhaps you should say
that a function is a map, rather than the other way around. Sometimes 
people say a mapping instead of a map. They may also use the word as a 
verb, saying that they map things. Other languages use words meaning 
image or reflection for the same concept. And why not, a (geographical) 
map is an image that reflects reality.

programming-is-doing-math-ly y'rs

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    05-Nov-99
Time:    10:51:25

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



From aahz at netcom.com  Wed Nov 10 22:02:24 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 11 Nov 1999 03:02:24 GMT
Subject: Namespace question...
References: <7vqgib$l8v$1@nnrp1.deja.com> <80c3e4$vpq$1@nnrp1.deja.com>
Message-ID: <80dbk0$qiu$1@nntp5.atl.mindspring.net>

In article <80c3e4$vpq$1 at nnrp1.deja.com>,  <raptor_99_uk at my-deja.com> wrote:
>In article <7vqgib$l8v$1 at nnrp1.deja.com>,
>  oofoe at my-deja.com wrote:
>>
>> How can you access the global namespace in python 1.5.1? I'd like to get
>> a version string which is declared in the main script file. However,
>> each module also has a version string which overrides the original one.
>> I tried to access it as __main__.Version, similarly to $main::Version in
>> Perl, which is what I'm more familiar with. Of course, it didn't work.
>
>just declare Version to be global inside bar() as
>follows:
>
>def bar():
>    global Version
>    print "Version is: ", Version

Your response is wrong: it works within a module, but not inside a
module that has been imported into another script.  Eric Jacobs gave a
correct response (import __main__), but in general it's a better idea to
pass information in (interface specified) rather than having a module
make assumptions about what kind of script will import it.
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From jeremy at cnri.reston.va.us  Fri Nov  5 16:15:51 1999
From: jeremy at cnri.reston.va.us (Jeremy Hylton)
Date: Fri, 5 Nov 1999 16:15:51 -0500 (EST)
Subject: IPC8: SouthWest Airlines Internet specials
In-Reply-To: <14371.16525.555599.961527@dolphin.mojam.com>
References: <Pine.WNT.4.04.9911051036270.260-100000@rigoletto.ski.org>
	<14371.14313.700685.807764@goon.cnri.reston.va.us>
	<14371.16525.555599.961527@dolphin.mojam.com>
Message-ID: <14371.18695.714122.631868@goon.cnri.reston.va.us>

>>>>> "SM" == Skip Montanaro <skip at mojam.com> writes:

  Jeremy> Everyone planning to fly to the conference should consider
  Jeremy> flying through BWI.  I have flown from BWI to San Jose and
  Jeremy> San Francisco a few times this year, each time for about
  Jeremy> $200.  Flights through BWI are often cheaper than flights
  Jeremy> from Dulles or National.

  SM> If you're on a budget and might like to get away without renting
  SM> a car, don't forget to figure in the cost of ground transport
  SM> into the total equation.  For IPC1 I flew into BWI and found the
  SM> metro didn't go there.  Not real convenient.  At least with
  SM> National you've got a Metro station nearby.

Good point!  It is possible to take public transportation, but it
isn't easy.  I just talked to a woman at the hotel about how to get
there from BWI and they recommended a taxi, with an estimated cost of
$50-$60.  She also mentioned a company called "1-800-BLUE-VAN" that
might offer a shuttle bus to the hotel; she was vague on details, but
it sounds like we have at least a phone number.

We also discussed two possibilities using public transportation.

You can take the commuter rail from BWI to Union Station in DC, and
get on the DC Metro (subway) from there.  The train ticket costs $5,
and the ride is about 20 minutes; check the schedule for departure
times.  There is a shuttle bus at the airport between the terminal
and the rail station.

If you took a cab to the hotel from Union Station, it would probably
cost around $15.

You could also take the Metro to the Rossyln stop, which is
two-and-a-half blocks from the hotel.  If you get there before 7:30pm,
the hotel runs a shuttle service that will take you the last leg.

There are Web pages for the commuter rail --
    http://www.mtamaryland.com/marc/marc_fare.htm
-- and the subway --
    http://www.wmata.com/

  SM> All that said, I don't know how convenient the conference
  SM> location is to public transport.  Jeremy, Guido, and other
  SM> locals (yikes! Guido's a local now?) can probably give a summary
  SM> of that sort of thing.

I hope that helps.  I don't know the area around the hotel, though;
perhaps some other local can fill in details if you've got more
questions.

I think the local travel between Dulles and National airports is a lot
simpler, although there is no public transportation that reaches
Dulles.  I'm fairly certain there are shuttle buses that get into DC
for a reasonable price, and a cab would be fairly inexpensive.

Jeremy




From emile at fenx.com  Sun Nov 28 12:19:12 1999
From: emile at fenx.com (Emile van Sebille)
Date: Sun, 28 Nov 1999 09:19:12 -0800
Subject: Binary vs non-binary
References: <19991127103504.A1536@stopcontact.palga.uucp> <slrn8414hl.eqn.quinn@riyal.ugcs.caltech.edu>
Message-ID: <08ef01bf39c4$b3ff75e0$01ffffc0@worldnet.att.net>

Also, in DOS OS's, 1Ax will be interpreted as EOF in non-Binary 
mode, and you may not get the whole file as a result.

-- 

Emile van Sebille
emile at fenx.com
-------------------






From gherman at darwin.in-berlin.de  Thu Nov  4 07:01:18 1999
From: gherman at darwin.in-berlin.de (Dinu C. Gherman)
Date: Thu, 04 Nov 1999 13:01:18 +0100
Subject: return boolean from functions
References: <7vrq9b$hmr$1@nnrp1.deja.com>
Message-ID: <3821758E.5080EAA4@darwin.in-berlin.de>

a_olme at my-deja.com wrote:
> 
> Is it possible to return booelan from functions other than 
> return 1==1 for true or return 1==2 for false?

There is no such thing like Booleans in Python. What you 
describe is simply the use of Python integers as Booleans,
in this case 1 and 0, respectively.

Real-booleans-are-not-bad-but-few-people-advocate-them'ly,

Dinu

-- 
Dinu C. Gherman
................................................................
Food for Echelon: Delta Force, SEAL, virtual, WASS, WID, Dolch,
secure shell, screws, Black-Ops, O/S, Area51, SABC, basement, 
ISWG, $@, data-haven, NSDD, black-bag, rack, TEMPEST, Goodwin, 
rebels, ID, MD5, IDEA, garbage, market, beef, Stego, ISAF, NARF, 
Manfurov, Kvashnin, Marx, Abdurahmon, snullen, Pseudonyms, MITM, 
Gray Data, VLSI, Leitrim... -- Visit http://www.hacktivism.org



From python-list at teleo.net  Thu Nov 11 11:10:18 1999
From: python-list at teleo.net (Patrick Phalen)
Date: Thu, 11 Nov 1999 08:10:18 -0800
Subject: How to perform syntax check in Python
In-Reply-To: <Pine.SOL.3.96.991111094750.5663A-100000@sundial>
References: <Pine.SOL.3.96.991111094750.5663A-100000@sundial>
Message-ID: <99111108162301.02136@quadra.teleo.net>

[Moshe Zadka, on Wed, 10 Nov 1999]:

:: Ummm....what planet are you from?
:: 
:: Python has no notion of declaring variables at compile time, only at
:: run-time. That means program flow can impact what name are "alive". This
:: means Python has to solve halting-problem-equivalent problems in order to
:: check for undeclared variables. I don't know what state HP is in, but if
:: you know how to solve the halting problem, you should patent it, and build
:: a brand new company. Guido has chosen not to release that technology into
:: Python for the next 5 years.

You seem tense, Moshe. Have you been drinking too much coffee?
Remember to take a break from the computer once in awhile. Go outside.
Look at the sky. Breathe. Relax.



From tekhir at yahoo.com  Mon Nov 15 00:28:28 1999
From: tekhir at yahoo.com (Jeremy Misavage)
Date: Mon, 15 Nov 1999 00:28:28 -0500
Subject: Setting up Python for IIS/PWS
References: <382F4A22.514281A8@yahoo.com>
Message-ID: <382F99FC.DEE9E08E@yahoo.com>

Oops. The <path to python> had spaces in in it and I just forgot to put quotes
around the path.

-----

 .py :REG_SZ: c:\<path to python>\python.exe -u %s %s



From evan at tokenexchange.com  Thu Nov  4 14:26:15 1999
From: evan at tokenexchange.com (Evan Simpson)
Date: Thu, 4 Nov 1999 13:26:15 -0600
Subject: Zope comparisons? (ASP, PHP, etc)
References: <3821C5F6.84543AF2@yahoo.com>
Message-ID: <06lU3.259$Ae3.186544@news-dal.corridex.com>

There has been a lot of discussion lately on the Zope mailing list
(zope at zope.org) about making this sort of information more readily available
on the website, but here's a brief intro:

Zope has several parts, which work well together, but are all optional:

1.  ZServer: A Medusa-based multi-threaded asynchronous server which accepts
HTTP, FTP, and monitor (special Python client) connections.  Usually, it
passes requests from these connections to...

2. ZPublisher:  An ORB which converts HTTP, FTP, XML-RPC, WebDAV, etc.
requests into traversal of an object hierarchy and returns the result of
rendering the target object.  These objects are normally stored in...

3. ZODB: A versioning, transaction-based object store with automatic object
persistence and conflict handling.  ZODB stores an object hierarchy as a
single file (2GB limit being worked on) and appends new versions of objects
when their request transaction is comitted.  This allows extensive undo, and
requires occasional packing to eliminate old, unwanted versions.  Most HTML
is generated by standard Zope objects using...

4. DTML: An embedded language similar to ASP and PHP, except that
expressions are written in Python and have access to the full power of the
rest of Zope, including...

5. Acquisition: A feature of standard Zope objects which allows them to
change their behavior automatically based on the context in which they are
called.  Hard to describe, but very powerful, as are...

6. Products: New objects or collections of objects can be added to Zope by
simply installing a Python package in the Products subdirectory and
resetting the server.  Current Products include database adapters, threaded
discussions, virtual host managers, and many others.  Of course, if you
don't want to learn Python you can always create...

7. ZClasses: Object classes defined through the web management interface by
assembling standard Zope parts.  These can provide a wide range of
customized objects.

I haven't really gotten into the user/role/permission security model, or the
extensive web management interface through which you typically interact with
Zope.  I may have forgotten other bits, too.  There's a lot there :-)

Charlie Hubbard <charliehubbard76 at yahoo.com> wrote in message
news:3821C5F6.84543AF2 at yahoo.com...
> I'm looking for some comparison info about zope.  Why zope?
> How is it different, the same, or better than ASP, PHP, or
> DHTML?  I've looked over some different docs on zope.org,
> but I wasn't able to find any info about what need zope fills.
> What can you do with zope that you can't do with ASP or PHP?
> It seems zope is based off of an object model where PHP and
> ASP are simplying server side scripting languages.  Does zope
> fill that need?  Is it in competition with this languages?  Would
> there be any reason to use Zope with ASP or PHP?






From endgamer71 at hotmail.com  Tue Nov  9 08:52:45 1999
From: endgamer71 at hotmail.com (Endgamer)
Date: Tue, 09 Nov 1999 13:52:45 GMT
Subject: Newbie question: Input
References: <3826b693.3888197@212.225.36.99> <rBAV3.9530$Xd2.521203@typhoon.nyroc.rr.com>
Message-ID: <382821d7.96897763@212.225.36.99>

On Mon, 8 Nov 1999 08:55:07 -0500, "Darrell" <news at dorb.com> wrote:

>From Python Library Reference:
>
>raw_input ([prompt])
>If the prompt argument is present, it is written to standard output without
>a trailing newline. The function then reads a line from input, converts it
>to a string (stripping a trailing newline), and returns that. When EOF is
>read, EOFError is raised. Example:
>
>>>> s = raw_input('--> ')
>--> Monty Python's Flying Circus
>>>> s
>"Monty Python's Flying Circus"
>>>>
>
>If the readline module was loaded, then raw_input() will use it to provide
>elaborate line editing and history features.

Cheers, Darrell. This is just what I was looking for!



Endgamer
Spider Warrior

(Endgamer71 at spamtrap.hotmail.com)
#####################################
Not a quote, just an on-the-spot observation:
You know you're a geek when your PC costs exactly
15 times as much as your stereo. And gives better
sound quality.<sigh>
                              Endgamer
#####################################



From fredrik at pythonware.com  Wed Nov 24 04:23:27 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 24 Nov 1999 10:23:27 +0100
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <8666ysfta9.fsf@ix.netcom.com>
Message-ID: <012e01bf365d$912fa360$f29b12c2@secret.pythonware.com>

Brad Knotwell <knotwell at ix.netcom.com> wrote:
> Similarly, the only thing I've found missing from Python is C's
> ternary operator, it would be nice to be able to do something like 
> the following:
> 
> def max(x,y): return ((x>y) ? x : y)

now that was one lousy example ;-)

Python 1.5.2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> max(1, 2)
2

</F>




From phd at phd.russ.ru  Sat Nov  6 10:32:30 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Sat, 6 Nov 1999 15:32:30 +0000 (GMT)
Subject: Get parameter by name
In-Reply-To: <3823FD98.8424037F@python.net>
Message-ID: <Pine.LNX.4.20.9911061532010.20586-100000@fep132.fep.ru>

On Sat, 6 Nov 1999, Thomas A. Bryan wrote:
> I'm not sure whether I understand the question.  Is this what you're 
> trying to figure out?
> 
> >>> ka = 1 
> >>> x = 'ka'
> >>> eval(x)
> 1
> >>> y = eval(x)
> >>> y
> 1
> 
> If not, please repost.

   Yes, but it is better to use locals() or vars() function for this...

Oleg.
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From dfan at harmonixmusic.com  Tue Nov 30 09:26:11 1999
From: dfan at harmonixmusic.com (Dan Schmidt)
Date: 30 Nov 1999 09:26:11 -0500
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk>
Message-ID: <wkiu2kqcn0.fsf@turangalila.harmonixmusic.com>

Mark Carroll <markc at chiark.greenend.org.uk> writes:

| Basically, I'd be really interested to learn: what _don't_ people
| like about Python? Every language has things that many people
| complain about, however reasonably, or would like to have been
| different, and I'd be interested to know what Python's are so I can
| get an idea if any are going to annoy me, rather than me being
| confused then frustrated then disappointed about them as I discover
| them further in. For instance, I'm sure plenty of people would hate
| Modula-3's lack of implicit casting, need to define everything
| before using it, Pascal-style uppercase keywords, etc. Are there any
| ways that Python makes programming slightly more awkward than people
| would have liked?

The main thing I find frustrating is that lambdas are not very
powerful.  Much of the time that I would use a simple 'map {}' in
Perl, I find that I have to write a separate function, or iterate by
hand.  My understanding, though, is that there's no plan to make
lambdas more powerful, which is a bit disappointing to me.

Michael Hudson's bytecodehacks do ameliorate some of my problems with
trying to write in a functional style.  I hope they make it into the
standard library at some point.

-- 
Dan Schmidt | http://www.dfan.org



From paul at digicool.com  Sat Nov 13 08:34:11 1999
From: paul at digicool.com (Paul Everitt)
Date: Sat, 13 Nov 1999 08:34:11 -0500
Subject: Production Web Deployment?
References: <NBBBKMMANMGEEBHPJJECGEGIFJAA.dirkl@home.com>
		<3826C69F.B4E26B89@digicool.com> <14379.18122.156471.644349@ghost.bbn.com>
Message-ID: <382D68D3.46CEFD61@digicool.com>

Tim O'Malley wrote:
>   Is Zope.org using Apache?
>   There is a prototype (alpha) Apache module which embeds the
>   functionality of PCGI into Apache.  If this worked for you, then you
>   would maintain compatibility with PCGI.

Howdy Tim, thanks for the note.  Yes, we know about Andrew's mod_pcgi. 
Our ultimate goal is to leverage the active development of another
group's webserver integration.

FastCGI gives us a three-year-old Apache module, plus commercial and
supported versions for Netscape and IIS (and Apache).  Additionally,
FastCGI has support for TCP sockets in addition to domain sockets, which
helps deal with file permission errors when configuring it.

The major reason we ditched FastCGI and came up with PCGI in the first
place (lack of threading in the library meant a person could telnet into
your application and lock the app) is no longer pertinent with the new
Zope2 architecture.  This has us hoping to get FastCGI back in as a
viable choice.  Seems nobody in the Zope community is eager to maintain
an Apache module. :^)

--Paul



From david at forestfield.co.uk  Sat Nov 13 13:25:00 1999
From: david at forestfield.co.uk (David F Hughes)
Date: Sat, 13 Nov 1999 18:25 +0000 (GMT Standard Time)
Subject: [HELP] TKinter could not run in Python!
References: <80i63e$7t4$1@news.seed.net.tw>
Message-ID: <memo.19991113182523.24605A@forestfield.cix.co.uk>

In article <80i63e$7t4$1 at news.seed.net.tw>, wuhy at tcts.seed.net.tw (Mark 
Wu) wrote:

> TclError: Can't find a usable init.tcl in the following directories:
>     {d:\progra~1\tcl\lib\tcl8.1} {} ./lib/tcl8.0 
> D:/Python/tcl8.0/library
> D:/Pyt
> hon/Tools/library
> 
  <Snip>

> 
> This probably means that Tcl wasn't installed properly.


As another Python newbie, I have just spent a couple of hours grappling 
with the same problem. I tracked it down to copies of Tcl80.dll and 
Tk80.dll that had gotten installed in [Windir]\system. Removing them 
solved the problem. 

It was also necessary to tell Python where to find the real Tcl 
installation by adding C:\Progra~1\tcl\bin to the PATH command in 
autoexec.bat - I think this would have been done automatically during the 
Python install, but it had been confused by the above.

I believe that the dll's had been put into the \system directory by a 
(rather garish)Tcl GUI builder called SpecTcl (from SUN microsystems dated 
September 1997) which I had installed to evaluate but decided not to use - 
well anyway, SpecTcl stopped working after removing them!

Regards,
David Hughes
Forestfield Software Ltd.
Horsham, UK.
[david at forestfield.co.uk] 



From gang.li at compuware.com  Thu Nov  4 16:21:14 1999
From: gang.li at compuware.com (Gang Li)
Date: Thu, 4 Nov 1999 14:21:14 -0700
Subject: input parameters dependcy?
Message-ID: <3821e773@199.186.16.51>

How can I know which inpute parameters will be used in a function, before
call the function.  e.g. If I have a function

def foo(self, var1, var2):
    a = self.left
    b = var1+var2
    c = a + self.top
    return  a,b,c

module test1 as following

g1 = [3]
g2 = [4]

class CL:
    __init__(self, l, t):
        self.left = l
        self.top = t

cl = CL(100, 20)

tt = foo(cl, g1, g2)

Before the call foo(cl, g1, g2), from the code of foo, I know cl.top,
cl.left, g1 and g2 will be used.  How can I detect which variables will be
used give the call condition foo(cl,g1,g2) before call it?





From I at DONT.LIKE.SPAM  Mon Nov 15 21:53:51 1999
From: I at DONT.LIKE.SPAM (David S.)
Date: Mon, 15 Nov 1999 18:53:51 -0800
Subject: Daemonic threads
Message-ID: <3830C73F.C8A7CA62@DONT.LIKE.SPAM>

I'm writing a CGI script that will launch a long-running process in the
background, and return some HTML page to the browser.  Now, I know how
to do that with 'fork()', but it seems to me that I should be able to do
the same thing with a daemonic thread from the 'threading' module.

I can't get that to work, though,

	...
	xt = threading.Thread(target=_threadExec, args=(save, mail, None))
	xt.setDaemon(1)
	xt.start()
	...

and the script exits with no thread running.  Non-daemonic threads seem
to work as I expect them to.  What don't I understand about daemonic
threads?  (Python 1.5.2 on Linux 2.2.7, if its relevant.)

David S.
davids@
...	bdgp.org



From asdkzjlaskld at oe80938.org  Mon Nov  1 06:36:06 1999
From: asdkzjlaskld at oe80938.org (asdkzjlaskld at oe80938.org)
Date: Mon, 01 Nov 1999 11:36:06 GMT
Subject: *Inkjetz 4 you
Message-ID: <9LX7HF6X.ICIZD40N@oe80938.org>

				The Inkjet Company has the following inkjet specials:
				All Prices Include Free Shipping and Free Sales Tax





EPSON

Epson -  $8.00 - Black  or 10 for $65.00 - All Models*
Epson -  $12.00 - Color or 10 for $100.00 - All Models*
*Epson - $16.00 - Black or 10 for $135.00 - 900/3000
*Epson - $19.00 - Color or 10 for $165.00 - 750/900/1200/3000


CANON

Canon BCI-21B - $5.00 - Black BJC-2500/4000/4400/5000
Canon BCI-21C - $8.00 - Color BJC-2500/4000/4400/5000
Canon BC-20 - $18.00 - Black BJC-2500/4000/4300/4400/
Canon BC-02 - $16.50 - Black BJC-210/230/BJ-200
Canon BC-05 - $28.00 - Tri-Color BJC-210/240
Canon BX-3 - $28.00 - Black B-540/550/640

(See Below For Apple & H.P. Products)


				
				The Inkjet Company

				1-310-517-9999
				1-503-905-6652 Fax


APPLE

M8041G/C - $17.00 - Black Styler Writer 1500
M4609G/C - $30.00 - Tri-Color Styler Writer 1500
M3240G/A - $19.00 - Large "Double" Black Style Writer 2400
M3330G/A - $5.00 - Black Style Writer 2400
M3329G/A - $9.00 - Tri-Color Style Writer 2400 
M1949G/A - $3.75 - Cyan Color Style Writer Pro 
M1950G/A - $3.75 - Magenta Color Style Writer Pro
M1951G/A - $3.75 - Yellow Color Style Writer Pro
M1952G/A - $3.75 - Black High Capcity Color Style Writer Pro


HEWLETT PACKARD

51625A - $19.00 - Tri-Color Deskjet 500c/550 Series
51626A - $16.50 - Black Deskjet Series / OfficeJet 300 / HP Fax
51629A - $16.50 - Black Deskjet 600 Series / OfficeJet 500 Series
51649A - $19.00 - Tri-Color Deskjet 600 Series / OfficeJet 500 Series
51645A - $16.50 - Black Deskjet 700/800/1000/1600
51640A - $16.50 - Black Deskjet 1200 / CopyJet



					The Inkjet Company

					1-310-517-9999
					1-503-905-6652 Fax

				    Visa, Mastercard, Amex, Discover
				Fast Free Delivery - usually in two days or less!



From skip at mojam.com  Wed Nov  3 11:39:17 1999
From: skip at mojam.com (Skip Montanaro)
Date: Wed, 3 Nov 1999 10:39:17 -0600 (CST)
Subject: assert module search
In-Reply-To: <7vpmi4$1gp$1@nnrp1.deja.com>
References: <000601bf25be$0cdeb880$2e2d153f@tim>
	<381FEF10.3B27C05B@webone.com.au>
	<7vpmi4$1gp$1@nnrp1.deja.com>
Message-ID: <14368.25909.535147.527730@dolphin.mojam.com>

    mehta> which module does assert function belong to?  I tried to link to
    mehta> it through api docs but led me to exceptions module.

Assert is a keyword in the language, not a function.  Docs for it are at

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

Skip Montanaro | http://www.mojam.com/
skip at mojam.com | http://www.musi-cal.com/
847-971-7098   | Python: Programming the way Guido indented...




From aa8vb at yahoo.com  Tue Nov 16 07:42:51 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Tue, 16 Nov 1999 07:42:51 -0500
Subject: python-sound-server for PySol under unix
In-Reply-To: <XFMail.991116065000.markus.oberhumer@jk.uni-linz.ac.at>
References: <XFMail.991116065000.markus.oberhumer@jk.uni-linz.ac.at>
Message-ID: <19991116074251.B711753@vislab.epa.gov>

Markus F.X.J. Oberhumer:
 |I've written a prototype implementation of a simple
 |sound server for the next PySol release using a 

Sounds pretty cool. :-)

 |small Python UDP server which forks of calls
 |to wav players like sox and bplay, but the
 |performance is not suitable for general release.
 |
 |Basically I'd only need what the Win32 winsound 
 |module provides (play/loop/stop), though further
 |extensions could be interesting for other applications.
 |
 |So is there an official Python-way of playing 
 |high-quality wav sounds under Unix ? 
 |Using pipes, sockets, or daemons like ESD ?

------------------------------------------------------------------------------
rplay is one fairly portable sound server out there

     http://rplay.doit.org/

>From the README, it supports Linux, SunOS 4.1.x, and Solaris 2.x.  FreeBSD
works as well.  The README says that Irix and HPUX support are supported
but status isn't known.  Win32 isn't listed.

------------------------------------------------------------------------------
A Python SDL wrapper sounds interesting too.  Adds Win32 portability but
hits a few less UNICes:

       http://www.devolution.com/~slouken/SDL/

You probably have investigated it more than I have, but at first glance, it
looks like it might be well suited for scripted access (in that much of the
processing is off-loaded to C).

FWIW, I recall reading that MpegTV (www.mpegtv.com) uses SDL for
full-screen DGA playback under XFree86.

Good luck!

-- 
Randall Hopper
aa8vb at yahoo.com



From juping+ at pitt.edu  Wed Nov 10 09:24:18 1999
From: juping+ at pitt.edu (Juping  Jin)
Date: 10 Nov 1999 14:24:18 GMT
Subject: Call a function in DLL with string arguments?
Message-ID: <80bv6i$kp$1@usenet01.srv.cis.pitt.edu>

I wrote a DLL and want its functions being available from python. Its function
accept string arguments. From python, I use windll to load the module like:

	handle = windll.module("dllname")
	handle.fcnname(arguments)

Python gives following msg:

TypeError: illegal argument type for built-in operation

So I did some tests, wrote a small dll and found out that if the function
accept only numerical values (actually, only integer value produces expected
results), then python won't give above errors. For char or char * argument,
it always gave above msg. Did I do something wrong?

Thanks for your reading.

Juping



From nemeth at iqsoft.hu  Tue Nov 16 05:09:25 1999
From: nemeth at iqsoft.hu (Nemeth Miklos)
Date: Tue, 16 Nov 1999 11:09:25 +0100
Subject: Beginning programming with Python
References: <sng0ychgwt.fsf@motorola.com> <382B3A53.5816B2@callware.com> <382B3D79.74CD3E0F@email.sps.mot.com> <80p9ds$ppo$1@newshost.accu.uu.nl> <asteroidal-80psfo/INN-2.2.1/bat@broadway.news.is-europe.net>
Message-ID: <38312D55.A36EB56D@iqsoft.hu>

Janos Blazi wrote:

> I have tried to use Perl in my teaching. It has a very powerful but alas!
> not very simple syntax! To start with, my pupils used to forget the dollar
> signs and Perl does not like that. So this year I shall use Python. We shall
> see. Time will tell, if Python is really better. Maybe my pupils will forget
> the indenting this time? There is actually nothing else you can forget in
> Python.

I am quite new in Python programming, and I was a bit :-( by understanding that
Python does not have any (even optional) mechanism to enforce the usage of
predeclared variables and functions. The complete lack of compile-time
type-safety of Pythom may be a problem in large projects, which may be the most
serious technical (ie non-marketing) obstacle to Python's becoming a widespread
language like Java.
Let us see an example:
# Module t.py
def f():
  print "started"
  x = 128
  x.foo()

Let us import it in the Python interpreter:
>>> import t
>>>
You can see thet the interpreter compiled t.py to t.pyc (or t.pyo) but found no
error.
All typos (typing errors of variable and function names) will be detected ONLY
at runtime! Python is perfect language to write programs fast, but with the
cost of hordes of testers. However this may not be a big problem, because any
piece of software should be thoroughly tested, to make sure that there are no
semantic errors in the application. In the case of Python the applications
should also be tested for "syntactic" errors.
In Python (just like in Smalltalk or in Perl, or in Tcl) there are a lot of
things a programmer may forget.
I am on my way of becomming a Python programmer, but this should not prevent me
from seeing the pitfalls of the language.
A tutor must emphasize these gotchas of any programming language.
Perl has a facility called 'use strict': it is also very poor compared to C++
and Java strict static (ie compile time) system.
The only really large project implemented in Python I found is Zope. Zope is a
fantastic piece of software, and may be regarded as an evidence of my being
totally wrong ;-).

> I feel that it is the right language for teaching.

Anyway, Python is a great language for teaching, but why to learn a language,
which is good only for teachning?! This is the real question you (or the Python
community) should answer!

Miklos Nemeth
IQSOFT





From gmcm at hypernet.com  Wed Nov 10 13:30:21 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 10 Nov 1999 13:30:21 -0500
Subject: statics everywhere.  Why?
In-Reply-To: <011701bf2ba4$1a899b80$f29b12c2@secret.pythonware.com>
Message-ID: <1269863880-796781@hypernet.com>

Fredrik wrote:
> Arinte <shouldbe at message.com> wrote:
> > Embedding python in a c application, why do I need
> > to make the functions that python call and variables
> > for callbacks static?
> 
> beats me.  what happens if you don't
> make them static?

On Unix, it exposes all your internals (with unpredictable 
results) to anyone loading your .so. On Windows, this isn't a 
problem, and much of the time you can control this with linker 
/ loader options on Unix. But since C promises not to let 
anyone else know about a file static, it's both the easy and 
safe route to portability.

- Gordon



From a.eyre at optichrome.com  Tue Nov 30 09:05:06 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Tue, 30 Nov 1999 14:05:06 -0000
Subject: WinNT and DLL's
In-Reply-To: <ddd64s4pp498hb7qlrgnq6o8fdqilrr27d@4ax.com>
Message-ID: <001701bf3b3b$e7a39570$3acbd9c2@peridot.optichrome.com>

>   Is there a way that I can simply access the pre-compiled dll files?

Have a look at the calldll module: http://www.nightmare.com/software.html

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From eric at net.tamu.edu  Wed Nov  3 16:11:46 1999
From: eric at net.tamu.edu (Eric Nelson)
Date: Wed, 3 Nov 1999 15:11:46 -0600
Subject: 64 bit ints for SNMP counter64 variables
Message-ID: <7vq8eg$4k$1@news.tamu.edu>

Is there not a 'long long' data type in python? I've got 64 bit counters
coming from snmp variables that I need to work with....

How about a variable precision type module??

Eric
eric at net.tamu.edu





From greg.ewing at compaq.com  Wed Nov 24 09:33:03 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Thu, 25 Nov 1999 03:33:03 +1300
Subject: filter function question - list processing in Python
References: <81f18b$hor$1@nnrp1.deja.com>
Message-ID: <383BF71F.3E413E33@compaq.com>

lobozc at my-deja.com wrote:
> 
> How should I pass additional parameter in 'filter'? 'map' has them,
> filter doesn't - why? Say I have a list ['xxx.py','xxx.pyc'] and I want
> to have a function isExtension(x,'.py')
> [No, I'm not interested in using lambda here in general].

Why not? That's the shortest way to do it:

   def filterExt(items, ext):
      return filter(lambda x, ext=ext: isExtension(x,ext), items)

However, a more Pythonic way of writing that is

   def filterExt(items, ext):
      result = []
      for x in items:
         if isExtension(x, ext):
            result.append(x)
      return result

which is longer-winded but clearer. If you want something which
is both short-winded and clear, you might like to try my List
Comprehension syntax extension, which allows you to write

   def filterExt(items, ext): # NOT STANDARD PYTHON
      return [x for x in items if isExtension(x, ext)]

The List Comprehensions patch is available from my web page:
http://www.cosc.canterbury.ac.nz/~greg

Greg








 - which I can call with various
> extensions to filter the list in variuos ways - how am I supposed to do
> that? [No, I'm not interested in using lambda here in general].
> 
> In general, I note that Python has some standard functions operating on
> lists but they are curiously limited. It is a pity, since these are the
> functions which reduce line count significantly - as well as make
> programming safer. I wonder what are the chances of repeating with
> lists what Python did with regular expressions - copying the
> functionality from a well tested and popular product?
> 
> Mathematica has a set of list processing functions. As one can expect
> from an essentially functional language and a product where most
> operations are done by list processing - for the last 10 years - this
> set is well thought out, complete and very, very useful.
> 
> [For description see
> http://documents.wolfram.com/v4/frames/frames.html, press button 'built-
> in functions', on the leftmost panel 'programming', to the right of it
> press 'functional programming' - eventually look for functions like
> Map, Fold, Apply etc.].
> 
> Sent via Deja.com http://www.deja.com/
> Before you buy.



From fdrake at acm.org  Wed Nov 24 10:39:01 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Wed, 24 Nov 1999 10:39:01 -0500 (EST)
Subject: Asyncchat Module Documentation and some asyncchat based modules
In-Reply-To: <19991123175832.A26756@gothic.andrich.net>
References: <19991123175832.A26756@gothic.andrich.net>
Message-ID: <14396.1685.377777.533389@weyr.cnri.reston.va.us>

Oliver Andrich writes:
 > Hi,

Hi, Oliver!

 > after I spend most of the last weeks implementing some asyncchat
 > based clients for common used protocols (at least I use the quite
 > common ;-) and a asyncchat based POP3 server. I like to ask if
 > somebody already started to document this module for the library
 > reference, otherwise I am willing and happy to do it.  I am so
 > impressed by asyncore and asyncchat, that I like to do it. ;-)

  asyncore is documented in the latest release of the documentation; I 
don't think it's included in the most recent Windows installer.  It's
easy to update the installed documentation under Windows; just unpack
the ZIP file in the right place and viola!

 > And I am looking forward to do a first release of my package of
 > asnyc* based clients in a week or two. And my POP3 server (100%
 > pure Python) in about three weeks. I have to to do some code
 > cleanups, security checks and performance tuning, but at the moment
 > the POP3 server runs fine in my testbed (simulating the 16.000
 > users I have on my production machine) and it outperforms cucipop
 > without any problems.

  Cool stuff!

 > So, who do I have to contact to supply documentation for asynchat?

  Me!  You can send it to me at python-docs at python.org, along with the 
appropriate release form.  See http://www.python.org/1.5/patch.html
for information about the release forms; we need that so we can
incorporate and distribute your additions.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From cfelling at iae.nl  Sat Nov 27 16:56:06 1999
From: cfelling at iae.nl (Carel Fellinger)
Date: 27 Nov 1999 22:56:06 +0100
Subject: Newbie: raw_input and other input methods
References: <19991125.1481400@adam-aurag.penguinpowered.com> <81kfjc$1vg$1@vvs.superst.iae.nl> <383EC1A5.99A4FAD5@lemburg.com>
Message-ID: <81pk1m$2i6$1@vvs.superst.iae.nl>

M.-A. Lemburg <mal at lemburg.com> wrote:

> It even does the Right Thing for the Python prompt. Try this:
...
> The above adds a little color to your average Python prompt.

Well, I was to save this for a rainy day, but now that you generalized it
I can't help. Besides it isn't nice to use terminal specific codes.


--------------------------------------------------------------------------
import readline, os


class PromptFunc:
    '''bucket for function to be evaluated during prompting'''
    def __init__(self, func, *args, **keys):
	self.func, self.args, self.keys  = func, args, keys

    def __str__(self):
	return apply(self.func, self.args, self.keys)


class Tput:
    '''terminfo capability string to be evaluated during prompting
    embedded with readline specials so it won\'t take up any space'''
    def __init__(self, attribute):
	self.attribute = attribute

    def __str__(self):
	return '\001' + os.popen('tput ' + self.attribute).read() + '\002'


class Var:
    '''(environment) variable to be evaluated during prompting'''
    def __init__(self, var, dict=os.environ):
	self.var, self.dict = var, dict

    def __str__(self):
	return self.dict[self.var]


class Prompt:
    def __init__(self, prompt='>>>', **state):
	self.prompt = prompt
	#colors; unfortunately xterm doesn't seem to obey terminfo names
	self.black, self.red = Tput('setf 0'), Tput('setf 1')
	self.green, self.yellow = Tput('setf 2'), Tput('setf 3')
	self.blue, self.magenta = Tput('setf 4'), Tput('setf 5')
	self.cyan, self.white = Tput('setf 6'), Tput('setf 7')
	#background colors
	self.Back, self.Red = Tput('setb 0'), Tput('setb 1')
	self.Green, self.Yellow = Tput('setb 2'), Tput('setb 3')
	self.Blue, self.Magenta = Tput('setb 4'), Tput('setb 5')
	self.Cyan, self.White = Tput('setb 6'), Tput('setb 7')
	#video attributes
	self.bold, self.blink = Tput('bold'), Tput('blink')
	self.dim, self.reverse = Tput('dim'), Tput('rev')
	self.flash, self.normal = Tput('flash'), Tput('sgr0')
	#emphesize
	self.begin_standout, self.end_standout = Tput('smso'), Tput('rmso')
	self.begin_underline, self.end_underline = Tput('smul'), Tput('rmul')
	#alternate charset
	self.begin_alternate, self.end_alternate = Tput('smacs'), Tput('rmacs')
	#override whatever
	self.__dict__.update(state)

    def __str__(self):
	return self.prompt % self.__dict__



if __name__ == '__main__':
    prompt = Prompt('%(red)s%(cwd)s(%(nr)d) %(green)shallo daar:%(normal)s ')
    prompt.cwd = PromptFunc(os.getcwd)
    prompt.nr = 0
    while 1:
	prompt.nr = prompt.nr + 1
	print raw_input(prompt)

-- 
groetjes, carel



From robert.meegan at murl.com  Thu Nov 11 14:26:56 1999
From: robert.meegan at murl.com (Robert Meegan)
Date: Thu, 11 Nov 1999 19:26:56 GMT
Subject: MySQLdb Module Question
In-Reply-To: <19991111140723.A7271@dmcom.net>
References: <19991111140723.A7271@dmcom.net>
Message-ID: <19991111.19265600@robert.polytopic.com>

Wayne -

According to the MySQLdb documentation, all parameters *must* be 
supplied as keyword arguments.

For example:

	host = 'localhost'
	user = 'wayne'
	password = 'secret'
	db = 'mydb'

	MySQLdb.connect(host = host, user = user, passwd = passwd, db = db)


- Robert 

-- 
Robert Meegan

Murl.com
Making the Internet easier to use - Anyplace, Anytime, Free!



>>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<<

On 11/11/99, 1:07:24 PM, Wayne Topa <wtopa at mindspring.com> wrote 
regarding MySQLdb Module Question:


>   As a Python newbie, I am having a problem getting the
> MySQLdbmodule-0.1.1 working.  I was able to use the
> MySQLmodule-0.4.1 to access my DB's but thought that I should upgrade
> to the newer module.

>   I can't connect to the Mysql database using
> >>> import MySQLdb
> >>> MySQLdb.connect('',user, passwd, Db)
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
>       TypeError: too many arguments; expected 1, got 5

> >From what I can understand from the Docs, MySQLdb.connect takes up to
> 9 args, so the above error really confuses me, as does this.

> >>> MySQLdb.connect()
> Traceback (innermost last):
> File "<stdin>", line 1, in ?
> File "/usr/lib/python1.5/site-packages/MySQLdb.py", line 259, in
> __init__
>   self.db = apply(connect, (), kwargs)

> Maybe I don't have the correct Docs?

> Could someone give me an example of some of the syntax of this module,
> please?

> TIA

> Wayne
> --
> According to my calculations the problem doesn't exist.
> _______________________________________________________

> --
> http://www.python.org/mailman/listinfo/python-list






From Gaetan_Corneau at baan.com  Wed Nov 10 08:29:58 1999
From: Gaetan_Corneau at baan.com (Gaetan Corneau)
Date: Wed, 10 Nov 1999 08:29:58 -0500
Subject: .dll and .exe file version informaton
Message-ID: <816010E2456BD111A48700805FBBE2EEFDF793@ex-quebec-u1.baan.com>

> go-ahead-just-*try*-to-write-a-program-to-do-that<wink>-ly y'rs  - tim

If you really want to do it in a program, take a look at the following
functions in the help:
	GetFileVersionInfo
	VerQueryValue
	QueryField

You wil find all this stuff crappy, but it works.
______________________________________________________
   Gaetan Corneau
   Software Developer (Quality Assurance Team)
   BaaN  Supply Chain Solutions  
   E-mail: Gaetan_Corneau at baan.com        
   Compuserve: Gaetan_Corneau at compuserve.com  
   ICQ Number: 7395494             
   Tel: (418) 266-8252          
______________________________________________________
"Profanity is the one language all programmers know best"



From mlh at vier.idi.ntnu.no  Tue Nov 23 20:50:10 1999
From: mlh at vier.idi.ntnu.no (Magnus L. Hetland)
Date: 24 Nov 1999 02:50:10 +0100
Subject: Visitor design pattern for python objects?
References: <818512$ma9$1@nnrp1.deja.com>
Message-ID: <y0jaeo4irot.fsf@vier.idi.ntnu.no>

Nicky <nickysandhu at hotmail.com> writes:

> Is there a visitor type method that exists for all python objects?
> Visitor methods would be useful to allow an apply type functionality
> for objects.
> E.g.
>  base python object has accept method
>  def accept(this,visitor):
>   visitor.visit(this)
>  The visitor than would be called back with the particular type of
> object allowing the python object to delegate behavior to the visitor
> object. Is this already implemented or is there an alternative way of
> doing this?

Wouldn't every subclass have its own accept-method to avoid using the
superclass in the visit-method? (Maybe I have missed some
coercion-rules here...) In that case, each accept-method may have a
visit-method name of a different name, at least... (visitSimpleNode
etc.)

suspecting-that-i-really-did-misunderstand-that-coercion'ly yours,

Magnus

> 
> --Nicky

--

  Magnus          Echelon jamming noise:
  Lie             FBI CIA NSA Handgun Assault Bomb Drug Terrorism
  Hetland         Special Forces Delta Force AK47 Hillary Clinton 



From donn at oz.net  Sun Nov 28 01:45:26 1999
From: donn at oz.net (Donn Cave)
Date: 28 Nov 1999 06:45:26 GMT
Subject: fcntl/ioctl and O_NONBLOCK
References: <slrn8405je.q0.hildeb@stahlw06.stahl.bau.tu-bs.de> <slrn840205.sn4.hildeb@stahlw06.stahl.bau.tu-bs.de> <slrn83t5mv.bhk.hildeb@stahlw06.stahl.bau.tu-bs.de> <81ml5t$eer$0@216.39.151.169> <81p32a$m14$1@216.39.151.169>
Message-ID: <81qj26$be6$0@216.39.151.169>

Quoth hildeb at www.stahl.bau.tu-bs.de (Ralf Hildebrandt):
...
| Fine. My problem here is the following: I start a program (plot) and
| feed it with random input data until it crashes. Output is irrelevant
| to me, I need only know if it crashes or exit()'s.
|
| On the shell:
| % plot < t1 > /dev/null
| Segmentation fault.
|
| In my Python program, though: The program ran for a while and never
| returned. It didn't crash, it didn't exit, nothing. It just sat there
| status "sleep" (displayed top). It didn't consume cpu-time, either.
|
| I figured I should empty fromchild (since the command above generates
| 8MB of output) from time to time. Then I came across the select()
| problem. 
|
| This is also fixed now, but now the popen3 Object generates about 1 MB
| of output (which I read using select() and read()), and then it STILL
| doesn't return. No crash, no exit. And once more, in top it says
| "sleep" once more. 
| Any ideas what's going on there?

Yes - only a guess, but it bites everyone at least once when trying
to talk to another process via pipes.  After each write(), where your
Python program is writing to the ``stdin'' of the plot process, add
a flush(), i.e. stdin.flush().

Python file objects use C buffered I/O, and when output is a pipe
device it's block buffered.  This means your data doesn't go directly
to the pipe, but into a buffer, and the pipe doesn't see it until a
certain amount has accumulated.  The flush() forces all the buffered
data to be written immediately anyway.  You can also control this
by the buffer size parameter to the function that creates the file
object, or just avoid a file object altogether and write directly
to the pipe file descriptor.

	Donn Cave, donn at oz.net



From johndoe0521 at my-deja.com  Tue Nov  2 03:57:49 1999
From: johndoe0521 at my-deja.com (johndoe0521 at my-deja.com)
Date: Tue, 02 Nov 1999 08:57:49 GMT
Subject: About RedHat 6.1 install program
Message-ID: <7vm92d$i1e$1@nnrp1.deja.com>

I think this might be python's problem.
I copy all the files from the CD of RedHat 6.1 to the harddisk, and
make under /misc/src/anaconda, and some errors occur. Should it happen?
And if I make some changes of the install source, how can I put the
new install program into boot image? Thanks!


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



From mhammond at skippinet.com.au  Tue Nov  2 16:41:37 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 3 Nov 1999 08:41:37 +1100
Subject: Assorted Questions regarding the future of Python...
In-Reply-To: <m3bt9cgbwy.fsf@phaedrus.zen.xom>
Message-ID: <004201bf257b$0b0c64f0$0501a8c0@bobcat>

> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Rob Hodges
>
> Unless something has changed dramatically since I last used windoze,
> VB is not compiled (in the sense of native machine code) any more
than
> a frozen python program; rather, VB "executables" have a small
header
> of native code that launches the vbrunxxx.dll (interpreter) to run
the
> bytecode that makes up the rest of the file.  Which is more or less

Actually, things have changed.  The latest version of VB has an option
to compile to native machine code.  This feature was created more from
a "check-box marketting" POV than anything else (ie, Delphi did it, so
VB decided it must too!)

Mark.




From cfelling at iae.nl  Wed Nov 24 13:47:22 1999
From: cfelling at iae.nl (Carel Fellinger)
Date: 24 Nov 1999 19:47:22 +0100
Subject: automatic "lootjes trekken" [ was Re: automatic "lotjes trekken"]
References: <81cqnv$2ip$1@vvs.superst.iae.nl> <81e3s2$j14$2@newshost.accu.uu.nl> <u7903p71c5.fsf_-_@ebi.ac.uk>
Message-ID: <81hbrq$1fh$1@vvs.superst.iae.nl>

Philip Lijnzaad <lijnzaad at ebi.ac.uk> wrote:
> On 23 Nov 1999 13:12:34 GMT, 
> "Martijn" == Martijn Faassen <m.faassen at vet.uu.nl> writes:

> Martijn> Not at all. everybody here speaks Dutch anyway. (it's a conspiracy!)

> well, nearly everybody. The real word is 'lootjes', not 'lotjes'. Sorry,
> couldn't resist. 

Well, the file at home *is* called lootjes-trekken.py, but somehow the
spellling got corrupted:)

OTOH, Dutch being a permissive language does allow for both variants!
"lotjes" being the generic form, "lootjes" the special case (from a
morphological point of view that is). But surely "lootjes trekken"
is the common way to refer to what we do at this time of the year.
-- 
groetjes, carel



From mgm at unpkhswm04.bscc.bls.com  Mon Nov 15 12:41:30 1999
From: mgm at unpkhswm04.bscc.bls.com (Mitchell Morris)
Date: 15 Nov 1999 17:41:30 GMT
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <80pe77$osc$1@hub.org>
Message-ID: <slrn830he9.n28.mgm@unpkhswm04.bscc.bls.com>

First, let me say you've done an excellent job ... working code is almost
always preferable to another discussion item. Good for you for doing
something I and many other have only complained about.

Second, every unattributed quote from this point on is my opinion ... you
should assume that every comment is followed by an IMHO or two until such
time as your blood pressure recedes.



In article <80pe77$osc$1 at hub.org>, T. Middletan wrote:
[snip]
><personal rant> I hate CPAN. I have always hated CPAN. I will always
>hate CPAN! All right, okay, in truth back when i was a PERL hacker i
>thought CPAN was the greatest, very convenient, it's true. But after
>the great light of Python descended upon me (actually it was kind of
>thrust upon me as a "strong suggestion", at a time when i had no
>interest in Python whatsoever --- thank D'Arcy of PyGreSQL fame!) my
>perspective changed. I realised PERL was amazingly painful to code,
>and even more painful to maintain --- why didn't i realise this
>before? And as the scales fell from my eyes i realised that CPAN was
>incredibly ugly and tedious to wade through, now that i thought about
>it. A sea of README files, and ugly filenames.

My personal take is that you never actually look at CPAN in a browser.  The
only useful interface to the Comprehensive Perl Archive is in CPAN.pm, which
is by default installed on your machine. That way you can search for useful
stuff without trying to figure out what (if any) disorganization scheme is in
use:

	$ perl -MCPAN -e shell

	cpan shell -- CPAN exploration and modules installation (v1.50)
	ReadLine support enabled

	cpan> i /DNS/
	Going to read /usr/local/CPAN/sources/authors/01mailrc.txt.gz
	CPAN: Compress::Zlib loaded ok
	Going to read /usr/local/CPAN/sources/modules/02packages.details.txt.gz
	Going to read /usr/local/CPAN/sources/modules/03modlist.data.gz
	Distribution    M/MA/MATTBM/DNS-ZoneFile-0.92.tar.gz
	Distribution    MFUHR/Net-DNS-0.12.tar.gz
	Module          DNS::ZoneFile   (M/MA/MATTBM/DNS-ZoneFile-0.92.tar.gz)
	Module          DNS::ZoneFile::Record (M/MA/MATTBM/DNS-ZoneFile-0.92.tar.gz)
	...

	cpan> readme DNS::ZoneFile
    ...

	cpan> wq DNS
	CPAN: CPAN::WAIT loaded ok
	Could not connect to the WAIT server at ls6.informatik.uni-dortmund.de port
	1404

	Trying your http proxy http://proxy.bscc.bls.com:8080/
	Use of uninitialized value at /usr/local/lib/perl5/site_perl/5.005/aix/Net/
	Cmd.pm line 308.
	Argument "" isn't numeric in eq at /usr/local/lib/perl5/site_perl/5.005/WAI
	T/Client.pm line 143.
	No greeting from server
	Use of uninitialized value at /usr/local/lib/perl5/site_perl/5.005/aix/Net/
	Cmd.pm line 308.
	Argument "" isn't numeric in eq at /usr/local/lib/perl5/site_perl/5.005/WAI
	T/Client.pm line 112.
	No luck with your proxy either. Giving up
	No searching available!

	cpan> quit

My longstanding complaints about CPAN.pm aside (mostly having to do with
sooper-seekrit keyword searching capabilities), this is a phenomenal win.
Having the ability to perform some non-trivial searching/cooking on the local
site makes up for a great deal of ugliness module-naming-wise. It also helps
mitigate the effects of a dog-slow proxy between us corporate cogs and the
rest of the world and the general unwillingness of many tools to speak to a
proxy server at all.



The mechanism that CPAN.pm uses is a set of three compressed text files
downloaded from CPAN that describe the current state of the archive. These
files have a short lifetime (three days, if my memory serves me) before
they're refreshed from the official source.

Please note that I'm not volunteering to write this, given the other
collections of things I'm having to ignore my family to get accomplished. I
do promise, however, to admire the eventual resulting utility. I may even be
able to submit patches thereunto.



As an aside, the other thing about Perl that I like is perldoc. As much as I
use lynx, it still sucks for searching the Python library reference. At some
point I'll write ``pythondoc'' to extract the parts I care about, but until
then I'll just keep kvetching.



always-happy-to-see-other-people-working-ly y'rs
+Mitchell

-- 
Mitchell Morris

I am Pentium of Borg. Division is futile. You will be approximated.



From emuller at painewebber.com  Sat Nov 13 17:10:02 1999
From: emuller at painewebber.com (emuller at painewebber.com)
Date: Sat, 13 Nov 1999 22:10:02 GMT
Subject: making the Python case to my dept.
References: <9911051740130D.02530@william.jinsa.org> <801p3q$q01$1@nntp2.atl.mindspring.net> <3824723A.5DE35FE5@callware.com> <99110611340304.02970@quadra.teleo.net>
Message-ID: <80knjo$57n$1@nnrp1.deja.com>

I hit the hp e-speak site and everything was in java. I was just
wondering where you got this news from?

In article <99110611340304.02970 at quadra.teleo.net>,
  Patrick Phalen <python-list at teleo.net> wrote:
> [Ivan Van Laningham, on Sat, 06 Nov 1999]:
>
> :: Aahz Maruch wrote:
> :: > I believe that there are more jobs for good Python programmers
than
> :: > people with the skills to fill them.  My company is certainly
having
> :: > difficulties (jobs in Mountain View! ask me!), and having HP
Cupertino
> :: > advertising for 10-12 Python positions only increases the
difficulty.
>
> :: I really hope that Aahz is correct.  As most of you know, I'm
stuck out
> :: here in Utah, which is widely known to be significantly behind
*any*
> :: curve you want to name.
> ::
> :: If there's a curve coming to Utah, I sure hope it will be one with
jobs
> :: for Pythonistas, because there certainly aren't *any* here now.
>
> Look at it this way: the HP jobs all appear to center around
developing
> the E-speak technology. Python has been chosen as the official
> scripting language for E-speak. Novell has publicly endorsed E-speak.
If
> Novell decides to builds Netware hooks into E-speak,  they'll probably
> hire rafts of Pythonistas in Utah.  :)
>
>


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



From mlh at vier.idi.ntnu.no  Fri Nov  5 09:18:46 1999
From: mlh at vier.idi.ntnu.no (Magnus L. Hetland)
Date: 05 Nov 1999 15:18:46 +0100
Subject: Drawing (directed) graphs
References: <000a01bf2792$b83ff150$3acbd9c2@peridot.optichrome.com>
Message-ID: <y0jeme5owc9.fsf@vier.idi.ntnu.no>

"Adrian Eyre" <a.eyre at optichrome.com> writes:

> > Good thing there is a platform independent Python graphics library!
> 
> How can there be. Some OSs (Minix) don't support graphics at all...

Maybe you should check out the URL given before whining <wink>

Even on Minix you could easily generate jpegs, or postscript files or
whatever with Piddle. If you can't run an X server or something, that
only means that you can't see it... Or maybe you have a printer?

--

  Magnus          Echelon jamming noise:
  Lie             FBI CIA NSA Handgun Assault Bomb Drug Terrorism
  Hetland         Special Forces Delta Force AK47 Hillary Clinton 



From kbaldermann at entire-systems.com  Tue Nov 30 08:57:40 1999
From: kbaldermann at entire-systems.com (Klaus Baldermann)
Date: Tue, 30 Nov 1999 14:57:40 +0100
Subject: int -> IP
References: <tmnZ3.304$vR2.30239@typ12.deja.bcandid.com> <3diu2xnnh8.fsf@amarok.cnri.reston.va.us> <_AAZ3.478$vR2.44873@typ12.deja.bcandid.com> <3dg0y1nimy.fsf@amarok.cnri.reston.va.us>
Message-ID: <3843d7ff@194.120.211.23>

Andrew M. Kuchling wrote in message
<3dg0y1nimy.fsf at amarok.cnri.reston.va.us>...
>sbarron at twilight. (Scott Barron) writes:
>> The string is a representation of the IP address as a long int in network
>> byte order.  For example 16777343 would be 127.0.0.1.  So it can't be


>def transip2 (ipstr):
>    # Pack into a 4-byte integer
>    s = struct.pack('!L', string.atol(ipstr) )
>
>    # Unpack as 4 bytes
>    L = struct.unpack('!4B', s)
>
>    return `L[3]`+'.' + `L[2]`+'.'+`L[1]`+'.'+`L[0]`

PMFJI a bit late, but the last line looks a bit tedious to type :-)
packing it the other direction (little endian) enables us using string.join:

>>> from struct import pack, unpack
>>> from string import atol, join
>>> def transip3(long): return join(map(str,unpack('4B', pack('<L',
atol(long)))), '.')

>>> transip3('16777343')
'127.0.0.1'
>>> transip3('2227275982')
'206.136.193.132'
>>>

map-filter-and-reduce-remind-me-of-APL-ly yours
Klaus






From fredrik at pythonware.com  Tue Nov  2 04:47:39 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 2 Nov 1999 10:47:39 +0100
Subject: Assorted Questions regarding the future of Python...
References: <6D8A17398E28D3119F860090274DD7DB49833A@pces.cadlab.it>
Message-ID: <006201bf2517$50edf1f0$f29b12c2@secret.pythonware.com>

Alessandro Bottoni <Alessandro.Bottoni at think3.com> wrote:
> So, I wonder:

(just some quick answers here.  others will surely
add to this).

> 1) How about a Python compiler? Does anybody is working on such a project? I
> saw a lot of messages regarding this topic in the past, but which is the
> current status?

well, python has a compiler (which converts source
code to byte codes).

but I suppose you mean "compile to machine code," for
which the answer can be summarized as "it's a hard pro-
blem, but people are working on it".  check the news-
group archives for more stories (look for things like
python2c, viperi, etc.).  and check jpython, of course.

> 2) How about a Python Virtual Machine, "a la Java", as a possible
> alternative to a real compiler? Is there anything like that? Is anybody
> working on it?

well, python has a virtual machine (PYC files contain
bytecodes, not compressed source code.  just like
Java class files).

as far as I know, there are no visible projects working
on faster ways to execute that code, but that doesn't
mean that they don't exist ;-)

(and don't forget jpython, which can convert python
sources to Java class files...)

> 3) How about using the "Template & Wizard" approach with Python, inside IDLE
> or another IDE? Does anybody is working on such a project? Guido, do you
> think to go in this direction with IDLE?

fwiw, we're working on something like this for future
versions of pythonworks.  not your usual dumb code
generator stuff, though.  more info some other day.

> 4) How about a "standard" installation tool? Is there anything like
> InstallShield or "VB Package & Deployement tool" in our world?
> Does anybody is working on such a project?

here's one:
http://starship.python.net/crew/gmcm/install.html

(and pythonworks includes similar tools, of course)

</F>

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




From jima at aspectdv.com  Mon Nov 15 15:21:26 1999
From: jima at aspectdv.com (Jim Althoff)
Date: Mon, 15 Nov 1999 12:21:26 -0800
Subject: method overriding trick
In-Reply-To: <lK9X3.696$An3.9333@newsfeed.avtel.net>
References: <14378.62498.649973.150075@bitdiddle.cnri.reston.va.us>
Message-ID: <4.2.0.58.19991115122013.00b0e540@mail.aspectdv.com>

I've been doing the exact same thing
only using _Super instead of _Parent.
I have found it to be a helpful convention.

Jim



At 12:58 AM 11/13/99 -0800, Phil Mayes wrote:
>Jeremy Hylton wrote in message
><14378.62498.649973.150075 at bitdiddle.cnri.reston.va.us>...
> >Overriding methods of a base class is a little clunky in Python.  You
> >need to explicitly name the base class that implements the method, and
> >you call its as an unbound method, passing self as the first argument.
> >Example: Base.aMethod(self, arg)
>
>
>and suggested using
>
> >class Derived1(Base):
> >    super_aMethod = Base.aMethod
> >
> >    def aMethod(self, arg):
> >        return self.super_aMethod(arg) + 1
> >
> >...
> >It doesn't solve the must-name-the-base-class problem, but it helps
> >manage it better.  If you keep the override assignments at the top of
> >the class statement, then you only have one place to look in each
> >class when the inheritance hierarchy changes.
>
>
>I use a method that doesn't remove the clunkiness, but localises the
>hierarchy to one line:
>
>_Parent = basedlg.BaseDlg
>class AddressDlg(_Parent):
>     def __init__(self, param):
>         self.local = param
>         _Parent.__init__(self, 2345)
>
>By using the same name "_Parent" in all modules, less thinking is needed,
>and cut'n'paste coding (what, me?) works better too.
>--
>Phil Mayes    pmayes AT olivebr DOT com
>Olive Branch Software - home of Arranger
>http://www.olivebr.com/
>
>
>
>
>--
>http://www.python.org/mailman/listinfo/python-list





From x at x.x  Wed Nov 24 17:20:36 1999
From: x at x.x (Eric Jacobs)
Date: Wed, 24 Nov 1999 17:20:36 -0500
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <81g8po$reo$1@mach.vub.ac.be> <002901bf3663$e4b62df0$11e9efc2@normik.dk>
Message-ID: <383C64B4.56CF6AF9@x.x>

maxm wrote:
> 
> > Mark Carroll (markc at chiark.greenend.org.uk) wrote:
> > [snip]
> > : Basically, I'd be really interested to learn: what _don't_ people like
> > : about Python?
> > [snip]
> 
> I miss deeply the ability to write:
> 
> i += 3
> Text += "Some new text\n"

There's a pretty good reason for this one, I think. The two statements:

  Text += "Some new text\n"
  List.append(9)

look sort of analogous. Both seem to be taking an object and performing
a method on it. But strings are immutable; you can't perform any method
on it to change the string. The syntax

  Text = Text + "Some new text\n"

makes it clear that we're creating a new object out of an old one, and
assigning it to the variable "Text".

Sure, you could interpret += to mean the same thing, but I think it
would be less clear to those trying to learn the language. It's critical
to get people out of C mindsets in order to be able to understand
what people mean when they say "everything is a reference."



From gmcm at hypernet.com  Mon Nov 29 13:06:33 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Mon, 29 Nov 1999 13:06:33 -0500
Subject: performance of tuple-less Python?
In-Reply-To: <oqso1pdwe7.fsf@titan.progiciels-bpi.ca>
References: "Greg Wilson"'s message of "Mon, 29 Nov 1999 11:29:30 -0500"
Message-ID: <1268223702-30481186@hypernet.com>

Fran?ois Pinard wrote:

> "Greg Wilson" <gvwilson at nevex.com> writes:
> 
> > [...] the use of tuples (which I believe are conceptually
> > redundant) [...]
> 
> Tuples may be used to index within dictionaries, and I find this
> feature very useful.  Lists may not.  I do not know the
> implementation details, but would be tempted to have the
> prejudice (:-) that using lists to index would imply some severe
> performance penalty.

I'm guessing that not being able to find something you stuck 
in the dictionary qualifies as a performance penalty ;-).

To spell it out:
 1) hash based on id(list_object):
     d = {[1] = 'x'}
     d[[1]] -> KeyError
 2) hash based on value:
     lst = [1]
     d[lst] = 'x'
     lst.append(2)
     d[lst] -> KeyError




- Gordon



From skaller at maxtal.com.au  Tue Nov 30 08:38:45 1999
From: skaller at maxtal.com.au (skaller)
Date: Wed, 01 Dec 1999 00:38:45 +1100
Subject: Python Type-Inference based LINT.. (pylint.py)
References: <19991121180043.B3184@teapot.egroups.net> <81e2r7$j14$1@newshost.accu.uu.nl> <383BC3D2.98BCD1FF@compaq.com> <81ktpm$33g$1@newshost.accu.uu.nl>
Message-ID: <3843D365.8B530226@maxtal.com.au>

Martijn Faassen wrote:
>It's hard to say, actually, because if it
> *has* to generate error messages like the kind you mentioned, optional
> type annotations may not even be worth it..

	I think they're worth it ... the main problem
is syntax. It isn't clear what syntax to use to specify
an annotation, and it isn't clear how to name the types.
The 'natural' notation:

	x : int = 1

can be used in function parameter lists, but not so easily for
variables (since ':' clashes with it's other use as a suite introducer).
It's also not clear that:

	x : PyInstance = SomeClass()

is useful, the class of x is more interesting.

-- 
John Skaller, mailto:skaller at maxtal.com.au
10/1 Toxteth Rd Glebe NSW 2037 Australia
homepage: http://www.maxtal.com.au/~skaller
voice: 61-2-9660-0850



From mhammond at skippinet.com.au  Tue Nov  2 19:07:03 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 3 Nov 1999 11:07:03 +1100
Subject: Reading .lnk files
In-Reply-To: <004301bf2583$b47a7150$0501a8c0@bobcat>
Message-ID: <004801bf258f$5c19d0d0$0501a8c0@bobcat>

Ivan has mailed me that this does not work for him (dies with a big
crash!).  If anyone else has a spare moment, could you please try this
code and see if it works?  Whether it does or not, please also include
info on the OS you tried it on.  Please mail me personally with this
info, and I will summarize back to the list.

Thanks,

Mark.

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Mark Hammond
> Sent: Wednesday, 3 November 1999 9:44
> To: 'Ivan Van Laningham'; 'Python Mailing List'
> Subject: RE: Reading .lnk files
>
>
> Sure is.  For more details, see the COM help file on PyIShellLink.
>
> Mark.
>
> # DumpShortcut.py
> # A demo of how to dump a shortcut using Python.
>
> # This code dumps the shortcut specificed as sys.argv[1]
>
> from win32com.shell import shell
> import pythoncom
>
> def DumpIt(fileName):
> # Get the shell interface.
>     sh = pythoncom.CoCreateInstance(shell.CLSID_ShellLink, None, \
>       pythoncom.CLSCTX_INPROC_SERVER, shell.IID_IShellLink)
>
>     # Get an IPersist interface
>     persist = sh.QueryInterface(pythoncom.IID_IPersistFile)
>
>     persist.Load(fileName)
>
>     # Get the data
>     print "Path name:", sh.GetPath(shell.SLGP_SHORTPATH)[0]
>     print "Args:", sh.GetArguments()
>     print "Description:", sh.GetDescription()
>     print "Working directory:", sh.GetWorkingDirectory()
>
> if __name__=='__main__':
>     import sys
>     DumpIt(sys.argv[1])
>
> # demo of usage:
> C:\>DumpShortcut.py IDLE.lnk
> Path name: L:\src\PYTHON~1.2\PCbuild\pythonw.exe
> Args: l:\src\python-cvs\tools\idle\idle.pyw
> Description:
> Working directory: L:\src\python-cvs\tools\idle
>
>
> --
> http://www.python.org/mailman/listinfo/python-list
>




From tseaver at starbase.neosoft.com  Tue Nov 16 10:52:25 1999
From: tseaver at starbase.neosoft.com (Tres Seaver)
Date: 16 Nov 1999 15:52:25 GMT
Subject: Integrate Phyton in a (Delphi) App ....
References: <383157E6.48E8@rbg.informatik.tu-darmstadt.de>
Message-ID: <8688A5AB30F13563.5711B9AB415CE2CD.6D202BACED783D20@lp.airnews.net>

In article <383157E6.48E8 at rbg.informatik.tu-darmstadt.de>,
Andreas Koch  <akoch at rbg.informatik.tu-darmstadt.de> wrote:
>Hi , i plan writing a huge app , in Delphi (Win95).
>It will have lots of classes for the various game
>objects , and i currently think about writing this
>class tree in Phython , to allow users changing it
>and adding new stuff.
>
>(for short , i call the delphi app winapp )
>
>So :
>1-How much work is it to integrate Phyton into a win app ?
>  The app will hold the complete framework for user
>  interface , movement , graphics , the phyton classes need to
>  call and be called by my procedures
>2-How fast is Phyton ? Speed is essential , i will have 
>  at least 100 objects per frame that need to be called and
>  (usually) call my winapps "drawtile" function.
>3-Is it possible to restrict certain program parts
>  from using certain code parts ? For example , don't
>  allow while-loops in a user definable module ?
>4-Is it possible for my winapp to execute some of
>  the phyton scripts in single step mode ?
> 
>							thanks , Andreas

You need to check out the Python for Delphi components, at
http://www.multimania.com/~marat/delphi/python.htm.

Perhaps also look at my "Snake in the Temple" presentation,
http://www.palladion.com/HADP_Jan99

Best,

Tres.
-- 
---------------------------------------------------------------
Tres Seaver           tseaver at palladion.com       713-523-6582
Palladion Software    http://www.palladion.com



From fredrik at pythonware.com  Fri Nov 19 08:36:18 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 19 Nov 1999 14:36:18 +0100
Subject: Python Chinese
References: <19991119121544.5006.qmail@hotmail.com>
Message-ID: <002001bf3293$10a89c10$f29b12c2@secret.pythonware.com>

Ulf Engstr?m <b2blink at hotmail.com> wrote:
> Is there any way to use Python with Chinese and other asian chars?
> If not, are there any plans to make this compability?

python's string type is 8-bit clean and doesn't
care about encodings (you have to keep track
of that yourself).

and yes, python 1.6 will grow unicode support,
which also includes support for various codecs.

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
(news: just posted an errata for the 'first printing') -->





From maytagrepairman at yahoo.com  Tue Nov 16 16:42:51 1999
From: maytagrepairman at yahoo.com (maytagrepairman at yahoo.com)
Date: Tue, 16 Nov 1999 21:42:51 GMT
Subject: Need help dealing with spaces in NT paths.
References: <80pr76$jq5$1@nnrp1.deja.com> <11f733ec.96958819@usw-ex0101-007.remarq.com> <uogcu93dt.fsf@cs.uu.nl>
Message-ID: <80sj4q$l7t$1@nnrp1.deja.com>

Ok.  This works so far, but to complicate things further I need also to
pass an argument to the command.  Both the command and the argument are
full path names with spaces in them.  If I type something like this at
the DOS prompt:

"my_full_path_cmd_with_spaces" "my_full_path_arg_with_spaces"

it works.

But if I try to do an equivalent such as this in Python

# this next statement is to see what is passed to os.system()
    print ('\"%s\" \"%s\"' % (cmd,arg))

# this next statement is the call to os.system()
    os.system('\"%s\" \"%s\"' % (cmd,arg))

I get

The name specified is not recognized as an
internal or extenal command, operable program or batch file.

Other variations I've tried give the same results.  Any thoughts on
this one?


In article <uogcu93dt.fsf at cs.uu.nl>,
  piet at cs.uu.nl wrote:
> >>>>> wheineman <wheinemanNOwhSPAM at uconect.net.invalid> (W) writes:
>
> W> Put quotes around the complete path. For example:
> W> myExec = "C:\\Program Files\\Trash Can\\compactor.exe"
>
> You probably mean:
> myExec = "\"C:\\Program Files\\Trash Can\\compactor.exe\""
>
> or easier:
> myExec = r'"C:\Program Files\Trash Can\compactor.exe"'
> --
> Piet van Oostrum <piet at cs.uu.nl>
> URL: http://www.cs.uu.nl/~piet [PGP]
> Private email: Piet.van.Oostrum at gironet.nl
>


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



From b2blink at hotmail.com  Wed Nov 24 18:11:20 1999
From: b2blink at hotmail.com (Ulf Engstrøm)
Date: Wed, 24 Nov 1999 18:11:20 CET
Subject: SMTP?
Message-ID: <19991124171120.1623.qmail@hotmail.com>

I've build a little mailthingy based on the 11.9.2 SMTP Example  from Python 
Library Reference but when I use it I'll get an empty mail with no sender 
and no msg, eventhough I don't get any errors whatsoever...Do I have to 
change something with the headers? If yes, what?

Using the example as is (with the change of the servername) doesn't work, 
it'll come out as:

send: 'ehlo WKS0002.DEVELOPMENT.b2b-link.com\015\012'
reply: '250-nebb.no Hello WKS0002.DEVELOPMENT.b2b-link.com, ESMTP 
hello!\015\012
'
reply: '250-VRFY\015\012'
reply: '250-EXPN\015\012'
reply: '250-ETRN\015\012'
reply: '250 SIZE\015\012'
reply: retcode (250); Msg: nebb.no Hello WKS0002.DEVELOPMENT.b2b-link.com, 
ESMTP
hello!
VRFY
EXPN
ETRN
SIZE
send: 'mail FROM:<alexis> size=10\015\012'
reply: '250 <alexis>, Sender accepted\015\012'
reply: retcode (250); Msg: <alexis>, Sender accepted
send: 'rcpt TO:<ulf.engstroem at nebb.no>\015\012'
reply: '250 <ulf.engstroem at nebb.no>, Recipient ok\015\012'
reply: retcode (250); Msg: <ulf.engstroem at nebb.no>, Recipient ok
send: 'data \015\012'
reply: '354 Enter mail, end with <CRLF>.<CRLF>\015\012'
reply: retcode (354); Msg: Enter mail, end with <CRLF>.<CRLF>
data: (354, 'Enter mail, end with <CRLF>.<CRLF>')
send: 'Small msg\015\012'
send: '\015\012.\015\012'
reply: '250 Ok, message saved\015\012'
reply: retcode (250); Msg: Ok, message saved
data: (250, 'Ok, message saved')
send: 'quit \015\012'
reply: '221 See ya in cyberspace\015\012'
reply: retcode (221); Msg: See ya in cyberspace

Thanx a lot for all previous help :)
Best regards

Ulf Engstr?m

http://www.b2b-link.com
ulf.engstrom at b2b-link.com

______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com



From g_will at cyberus.ca  Mon Nov  8 09:02:47 1999
From: g_will at cyberus.ca (Gordon Williams)
Date: Mon, 08 Nov 1999 14:02:47 GMT
Subject: Fastest GUI toolkit?
Message-ID: <01bf29f1$cbfd5e60$5a41c3d1@Cyberus.cyberus.ca>

I have an application where I need to draw many lines fast and repeatedly
(an oscilloscope).  The number of lines ranges from about 1000 to 16000. 
Currently I have the application using Tkinter but I am finding it too
slow.  I am currently using W95 as an os, but would prefer to make the code
portable across other os if possible.  Right now speed is more important
than portability so I would be willing to give this up.

I would like to draw 1000 lines in 1/10 of a second on a P166 machine.  (Am
I way off the mark??)  I was hoping that someone would have tried different
packages and could give me pointers in the right direction.

Thanks,

Gordon Williams



From cgw at fnal.gov  Thu Nov 25 09:33:32 1999
From: cgw at fnal.gov (Charles G Waldman)
Date: Thu, 25 Nov 1999 15:33:32 +0100 (MET)
Subject: decompress a bunch of .gz files
In-Reply-To: <000501bf374e$116bafb0$3acbd9c2@peridot.optichrome.com>
References: <81jc8i$ilp$1@nnrp1.deja.com>
	<000501bf374e$116bafb0$3acbd9c2@peridot.optichrome.com>
Message-ID: <14397.18620.87516.687622@gargle.gargle.HOWL>

Adrian Eyre writes:
 > > I'm searching for a way to decompress all .gz
 > > files in a directory automatically.
 > 
 > Unix:
 > 
 > #!/bin/sh
 > for file in *.gz; do gunzip $file; done

Or just:
   
 gunzip *.gz





From greg.ewing at compaq.com  Wed Nov  3 04:24:34 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Wed, 03 Nov 1999 22:24:34 +1300
Subject: tkinter grid() question
References: <381F2D32.9E3A15CF@research.att.com>
Message-ID: <381FFF52.707400E5@compaq.com>

"Garrett G. Hodgson" wrote:
> 
> in the code below, i create a Text entry area with a Scrollbar.
> i want this pair to always use all of the available space.
> but when i resize the window, the stay the same size, neither
> expanding or shrinking.

You have to do two things: make the Text widget sticky on
all sides, and configure the weight of the row and column
containing the Text widget so that it sucks up all the
extra space.

It's simplest if the widgets only take up one cell of
the grid each, so I'll show you how to do it that way:

   self.body.grid(row = 0, column = 0, sticky = N+S+E+W)
   scrollbar.grid(row = 0, column = 1, sticky = N+S)
   self.rowconfigure(0, weight = 1)
   self.columnconfigure(0, weight = 1)

Since you're not putting anything else in the frame in
the code you showed, there's really no reason to have
anything spanning more than one row or column. If you
want to control the initial size of the Text widget,
use the width and height options when you create it.

If you're going to put other widgets in the frame as
well, you'll have to be careful how you configure the 
grid so that resizing the text widget properly doesn't  
have undesirable side effects on the layout of the other
widgets. 

Essentially, you'll have to pick one of the columns spanned
by the text widget and configure it's weight to be 1, but
not one that has another widget in it that you don't want
stretched as well. Or, instead of trying to put everything
in one frame, you may find it easier to use more subframes
so that you can control their layouts independently.

I can't give any more specific advice without knowing more 
about what you're trying to do.

Hope that helps,
Greg



From gmarry at erols.com  Tue Nov 30 16:16:17 1999
From: gmarry at erols.com (Garry Hodgson)
Date: Tue, 30 Nov 1999 16:16:17 -0500
Subject: bug in tkinter canvas event handling?
References: <38429BB0.1BFFB54B@sage.att.com> <3843C0A0.94990E64@home.com>
Message-ID: <38443EA1.923861B6@erols.com>

Doug Hellmann wrote:

> This behavior surprised me at first, too.  It's as though the first
> canvas object is grabbing the pointer so that all events go to it.  I
> got around the problem by looking for the x,y coords of the release
> event and then finding the object(s) under that point on the canvas.

well, some earlier code i had did its own hit detection.  but it
seemed silly to do it myself when the canvas oughta do it for me.
i've yet to peruse the source, per Fredrik's suggestion, so i don't
know whether Tk considers this bug or feature.  i know how i'd vote.

-- 
Garry Hodgson				And it's just a box of rain
Software Innovation Services		I don't know who put it there		
AT&T Labs				Believe it if you need it
garry at sage.att.com			Or leave it if you dare



From newsreporter at rciinfo.com  Mon Nov 29 03:02:59 1999
From: newsreporter at rciinfo.com (News Reporter)
Date: Mon, 29 Nov 1999 3:02:59 PM -0500
Subject: Finding Good Human Resource Candidates
Message-ID: <199911291502.YND4953@rciinfo.com>

The competition for finding good Human Resource candidates if fierce! However, there is an edge - and we want you to have it FREE!

BestJobsUSA.com will run all of your Human Resource category positions free. No catch. Nothing to buy. No obligation whatsoever.

At BestJobsUSA.com, we recognize our business is dependent upon getting results for our Human Resource clients. This is just another way to keep you aware of our services.

E-mail your Human Resource category positions to us at: mailto:RCI at BestJobsUSA.com. We'll have it up and running within 48 hours.

If this isn't your function please forward on to the appropriate person.

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

We also want to make you aware of our WebRunner resume retrieval technology.

WebRunner will deliver resumes that meet your specific job criteria quickly and cost effectively.

This service is available now for the low cost of $400.00 per search. Simply provide us with the with job description, or key words (competitors you would like to target
) and we do all the work.

For full information, contact Gregg Lowenstein at mailto:glowenstein at rcimedia.com  or telephone at (561)686-6800, ext. 314

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

This information has been e-mailed to keep you apprised of recruitment-advertising opportunities. If this information is not suitable for your area of responsibility or if you want to be removed from our distribution list, please e-mail mailto:nothankyou at rcimedia.com?subject=HR






From gmcm at hypernet.com  Wed Nov 24 15:01:34 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 24 Nov 1999 15:01:34 -0500
Subject: whitespace.py module
In-Reply-To: <y7b66yrdel7.fsf@agastya.eng.fore.com>
Message-ID: <1268648802-4911011@hypernet.com>

Rajesh Vaidheeswarran writes:
> 
> I'm new to the python language, and the newsgroup. So, if this
> newsgroup is not appropriate for this type of posting, please
> pardon my ignorance and notify me.
> 
> I have written up a module called `whitespace.py' available at
> <URL:http://www.dsmit.com/python/>
> 
> and a sample program `whitespace' which uses this module, also
> available at the same URL.
> 
> The module can be used to check a given file for whitespace
> problems as defined in the module's documentation (also available
> as a separate text document at
> <URL:http://www.dsmit.com/lisp/whitespace.txt>.
> 
> Please feel free use these, and also to send in your comments on
> either of these.

I had a quick look. You are taking the opposite approach from 
the recommended way of dealing with tabs. The 
recommended approach is to set your editor to always insert 
spaces. That way, someone else reading your code will 
always see sane indentation, no matter how they have 
configured their editor (or what the editor assumes). You'll also 
make email clients happy (since they almost always turn 
leading tabs into spaces).

It's also acceptable to use only tabs for indentation.

> I would like to know how to submit a module for inclusion into
> the standard python distribution, and would appreciate any
> pointers in that direction.

There's already a whitespace troubleshooter in the std lib - 
tabnanny.py (check out python -t ).

You can transform any legal style of indentation to any other 
legal style with tabcleaner.py in the System directory of the 
contrib ftp site.

- Gordon



From Jan.Walter at t-online.de  Mon Nov  1 14:15:32 1999
From: Jan.Walter at t-online.de (Jan Walter)
Date: Mon, 01 Nov 1999 20:15:32 +0100
Subject: ANNOUNCE: Free Python Scripts For Blender (export to free renderers)
Message-ID: <381DE6D4.6867@t-online.de>

Hi,

At http://www.q-bus.de/Blender you can find a web page from which you
can download several Python scripts to export to following renderers
within Blender (http://www.blender.nl):

Panorama  (http://www.gnu.org/software/panorama/panorama.html)
Povray    (http://www.povray.org)
Radiance (http://radsite.lbl.gov/radiance/HOME.html)
RenderMan (http://www.renderman.org, http://www.bmrt.org)

I try to compare this renderers and any comments are welcome ...

Cheers,

Jan



From ivanlan at callware.com  Wed Nov 10 16:54:24 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Wed, 10 Nov 1999 14:54:24 -0700
Subject: Popup menus in Tkinter
References: <199911081726.SAA25412@axil.hvision.nl> <80796f$f9h$1@nnrp1.deja.com> <x7u2mu6tsu.fsf@lindm.dm>
Message-ID: <3829E990.F233BE08@callware.com>

Hi All--

Dieter Maurer wrote:
> 
> dkaznadzey at my-deja.com writes on Mon, 08 Nov 1999 19:44:16 GMT:
> > What would be a good source of information on tk commands
> > available through Tkinter?
> I use "tclhelp", a program with comes with "tclX".
> It is a grafical documentation program for "tcl", "tclX"
> and "tk". Alternatively, you can use the online tcl/tk
> documentation (e.g. from scriptics).
> 
> These, of cause, document "tcl/tk" not "Tkinter".
> However, the mapping from "tk" to "Tkinter" is quite straight forward:
> 

John Grayson's forthcoming book on Tkinter programming with Python has a
good section on the mappings.  Lots of other good stuff, too.

<plugs-r-us>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours
----------------------------------------------



From aa8vb at yahoo.com  Tue Nov  2 07:22:02 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Tue, 2 Nov 1999 07:22:02 -0500
Subject: Tkinter canvas blow-up:  bbox(ALL) == None
In-Reply-To: <013c01bf2510$57457f20$f29b12c2@secret.pythonware.com>
References: <19991101102103.A134035786@vislab.epa.gov> <013c01bf2510$57457f20$f29b12c2@secret.pythonware.com>
Message-ID: <19991102072202.A134170008@vislab.epa.gov>

Fredrik Lundh:
 |Randall Hopper <aa8vb at yahoo.com> wrote:
 |>      Apparently if the bounds of the items in a Tk canvas widget exceed
 |> MAXINT, bbox(ALL) returns None.  The attached Python script demonstrates
 |> this.
 |> 
 |>      Intuitively it seems like this is a bug since the contents of the
 |> canvas are maintained in floating point; integer bounds shouldn't be
 |> involved, should they?
 |> 
 |>      Is this a Tk bug or a Tkinter bug?
 |
 |both.
 |
 |Tkinter uses _getints instead of _getdoubles to
 |convert the bounding box to a tuple...
 |
 |...but the reason you get None instead of an over-
 |flow error is that Tk returns an empty string in this
 |case (at least in 8.0.5).


Ok, thanks for the info.  So do you know if Tk uses ints internally as
well, or is its returning the empty string an artifact of how it was fed by
Tkinter?

Randall

-- 
Randall Hopper
aa8vb at yahoo.com



From ionel at psy.uva.nl  Sat Nov 13 16:56:35 1999
From: ionel at psy.uva.nl (Ionel Simionescu)
Date: Sat, 13 Nov 1999 22:56:35 +0100
Subject: Max Size of Array when using Numeric.py
References: <38286054.C8612EB5@edf.fr> <nygX3.2040$V4.64389@news1.frmt1.sfba.home.com>
Message-ID: <80kmnv$dmd@mail.psy.uva.nl>

Paul F. Dubois <dubois1 at llnl.gov> wrote in message
news:nygX3.2040$V4.64389 at news1.frmt1.sfba.home.com...
| There is no intrinsic limit on the size of NumPy arrays. You just ran out
of
| memory, I guess.

"just run out of memory" will happen when your memory manager cannot
allocate that much space of *contiguous* memory. This can have puzzling
effects when your heap is badly fragmented.

| You could write a C extension that created the array and did the reading
of
| each element and placing it into the array so that you only had the one
| copy.

While I see that arrays defined by extension modules can be guaranteed to
remain unique, the same can be also assured with Numeric, because arrays can
be written "in place".

However, when you need to change the type of an array, you must duplicate it
anyway, even if your arrays are defined in extensions.


Helka FOLCH <helka.folch at edf.fr> wrote in message
news:38286054.C8612EB5 at edf.fr...
> I've tried to create an array object whose shape is (5500,200) but I get
> a memory error. Is there a maximum size for array objects?
>
> Maybe there's a better way of creating the array than what I did :
> My program reads the array which is in another format from a file. It
> transforms the array to the right format and holds it in a variable(mv)
> as a string. Then I just do:
>
> m = array(eval(mv),Float)


If you afford to use an <eval> to get the actual values, then your string,
<mv>, should be really large, and it may be it the one which eats up your
memory, rather than the arrays.

Ways to really improve the speed of reading & converting the data can rely
on the struct standard module.

However, those are really useful when you must tackle heterogeneous data
series (i.e. data structures).

For homogeneous data series you can use Numeric's own methods:

a = fromstring(data, typecode)
and
data = a.tostring()

, where data is a binary string you can read from/write into a file, or,
even better, store in a database.

ionel










From Gaetan_Corneau at baan.com  Wed Nov  3 11:53:31 1999
From: Gaetan_Corneau at baan.com (Gaetan Corneau)
Date: Wed, 3 Nov 1999 11:53:31 -0500 
Subject: __LINE__?
Message-ID: <816010E2456BD111A48700805FBBE2EEFDF750@ex-quebec-u1.baan.com>

Is there an equivalent of the C __LINE__ macro in Python to get the
currentline number?

Obviously it is available, since exceptions report it.

Sorry if this is FAQ, I can't find it.

Thanks,
______________________________________________________
   Gaetan Corneau
   Software Developer (Quality Assurance Team)
   BaaN  Supply Chain Solutions  
   E-mail: Gaetan_Corneau at baan.com        
   Compuserve: Gaetan_Corneau at compuserve.com  
   ICQ Number: 7395494             
   Tel: (418) 266-8252          
______________________________________________________
"Profanity is the one language all programmers know best"




From TUpchurch at romac.com  Wed Nov 10 14:43:44 1999
From: TUpchurch at romac.com (Upchurch, Tory)
Date: Wed, 10 Nov 1999 14:43:44 -0500
Subject: Python Job opportunity
Message-ID: <F23572BD734ED211858C00104B216B95036B2C7B@rta1cn04.romac.com>

I am working with a company that is a leading Internet destination dedicated
to empowering consumer decision-making through the exchange of
user-generated information, knowledge and opinion. The site features more
than 600,000 people-powered ratings of more than 25,000 products, services
and topics in nearly 800 categories, as well as related online discussion
forums.  This company is based in Austin,TX.  This is a full time permanent
position with a very good pre-IPO company.  

This is a great opportunity for someone that wants to live in Austin,TX.
If you are interested in this position, please send me a copy of your resume
to tupchurch at romac.com.

This position would be responsible for working with a team to define
requirements and support products.  You will also be responsible for
designing and coding to specifications.  

The requirements are as follows:
Intermediate Python programming ability
CGI development

Desired skills:
Expert Python skills
Knowledge of RDB/SQL
Web based advertising system experience
Customer management experience


Tory Upchurch		
Romac International Inc.
Recruiter IT Division
Direct (512) 231-3604
Toll Free (888) 548-6514
Fax (512) 345-7736	
tupchurch at romac.com
http://www.romac.com 




From jcw at equi4.com  Fri Nov  5 06:45:17 1999
From: jcw at equi4.com (Jean-Claude Wippler)
Date: Fri, 05 Nov 1999 12:45:17 +0100
Subject: ANN: MkWrap 0.9 embedded database
Message-ID: <3822C348.58034399@equi4.com>

This is an announcement of a new release of MkWrap for Python at:
	http://www.equi4.com/metakit/python.html
Ready-to-run downloads for Solaris, Linux, Windows, and Mac are at:
	ftp://ftp.equi4.com/pub/mk/

To give you an impression of what MkWrap does, here's a Python script:
    import MkWrap
    db = MkWrap.Storage('phonebook.dat',1)
    vw = db.view("persons")
    vw = vw.Select(name='John').Sort(vw.date)
    for i in vw:
        print "Found:", i.name, i.phone, i.date

MetaKit packs a lot of power in a little package.  Some facts:
	- Failsafe: transacted commit, for data and structure changes
	- Efficient: compact auto-sizing storage, high scanning speeds
	- Flexible: select/sort on any field, relational / set operators
	- Adaptive: restructure on-the-fly, use with C++, Python, and Tcl
	- Portable: the same file can be used on Unix, Windows, and Mac
	- Embeddable: can link statically for stand-alone executables
	- Tiny: on Windows, the no-install Python extension DLL is 170 Kb

MetaKit is free for non-commercial use, there's a royalty-free source
license for commercial use.  The non-nonsense details are described at:
	http://www.equi4.com/metakit/licensing.html

Changes since 0.4  - MkWrap 0.9 is based on MetaKit 1.9d, and adds: 
 * stability, this version has taken quite a beating 
 * a file interface, read/write/pickle to/from memo fields 
 * a wrapper around any sequence, so you can join/groupby anything 
 * completely reorganized core-, platform-, and language-distributions
 * smaller, faster, some documentation, and some sample scripts 

-- Jean-Claude

________________________________________________________________________
Jean-Claude Wippler    MetaKit home page - http://www.equi4.com/metakit/
Equi4 Software         "Portable database software for a changing world"



From smst at quantisci.co.uk  Tue Nov 16 06:06:59 1999
From: smst at quantisci.co.uk (Steve M S Tregidgo)
Date: Tue, 16 Nov 1999 11:06:59 GMT
Subject: Memory not being released?
Message-ID: <199911161106.LAA13046@cobweb.quantisci.co.uk>

We have glibc 2.0.7 -- does anybody know if the problem with v2.1.1
mentioned below exists with v2.0.7?

mickut.at.iki.dot.fi at python.org wrote:
> 
> I had similar problem just a while ago, and Cyclops helped a lot.
> But it didn't solve the whole problem, as it was not python's
> fault, but a Linux library glibc was functioning improperly. If you
> are running RedHat 6.0 and/or have glibc version 2.1.1, you should
> upgrade asap. Upgrading to 2.1.2 the bloating dissappeared and
> everything has worked nicely ever since.
> 
> Antti Kuntsi
> 
> --
> ______________________________________________________________
> |O| ----------------------------------------------------------
> | | |Twisted mind? No, just bent in several strategic places.|
> `\| |___________________________________mickut at iki.fi/mickut/|
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list

Cheers,
Steve



From c.evans at clear.net.nz  Sat Nov 27 16:45:30 1999
From: c.evans at clear.net.nz (Carey Evans)
Date: 28 Nov 1999 10:45:30 +1300
Subject: SMTP - again ;)
References: <19991126080017.26482.qmail@hotmail.com>
Message-ID: <874se7a9s5.fsf@psyche.evansnet>

"Ulf Engstr?m" <b2blink at hotmail.com> writes:

> Anyway, something else came up, when I enter a name in the header:
> "To: Ulf Engstr?m" the mail gets to the dest as:
> 
> To: Ulf.Engstr?m at mb04.swip.net

The address in the To: field is required to be a real address (with an
@ in it), so most mail servers will rewrite it when they receive it.
You might get what you want if you put the name in a comment:

  To: (Ulf Engstr?m)

-- 
	 Carey Evans  http://home.clear.net.nz/pages/c.evans/

	       "This is where your sanity gives in..."



From kcazabon at home.com  Mon Nov 15 23:35:06 1999
From: kcazabon at home.com (Kevin Cazabon)
Date: Tue, 16 Nov 1999 04:35:06 GMT
Subject: win32api... another question:  GetEnvironmentVariable("PATH")
References: <01BF2F61.1D2B8DB0@exchange1.geometria.hu> <W8WX3.67146$Rx2.793309@news1.rdc1.bc.home.com> <OL%X3.4846$Xo.25269@news-server.bigpond.net.au>
Message-ID: <_95Y3.69699$Rx2.821068@news1.rdc1.bc.home.com>

Cool... at least I know that I'm not loosing my mind.  q:]

I'll try other routes for now.  Is there any way to use the Win32 API
"SetEnvironmentVariable" function as well?  I want to update some system
stuff on the fly...

BTW, it doesn't cause any errors, just returns None.

Kevin.

P.S.:  the rest of the win32 package is pretty slick... thanks!  q:]


Mark Hammond <mhammond at skippinet.com.au> wrote in message
news:OL%X3.4846$Xo.25269 at news-server.bigpond.net.au...
>
> Kevin Cazabon wrote in message ...
> >Interesting... it doesn't work on Win98 for me (The exact statement
> >below...).
>
> Eeek - you are correct - it doesnt :-(
>
> win32api implements this using "GetEnvironmentVariableW()", and returns a
> Unicode string.  This doesnt work on 95/98.  I dont know what I was
thinking
> :-(
>
> I think I should drop this back to an Ascii version, and return a normal
> PyString object.  That shouldnt break any code as Unicode objects are
pretty
> useless now and need to be "str()"d to be useful.
>
> Damn...
>
> Mark.
>
>
>





From tim_one at email.msn.com  Thu Nov 25 05:56:10 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 25 Nov 1999 05:56:10 -0500
Subject: Python complaints
In-Reply-To: <81itrf$eqc$1@newshost.accu.uu.nl>
Message-ID: <000401bf3733$af411100$be2d153f@tim>

[Aahz Maruch]
> ...
> In fact, I wish that Guido would disallow
>
>   if foo : bar()
>
> and force it to always be
>
>   if foo :
>     bar()

He would be more likely to force it to be

    if foo:      # he hates spaces before colons
        bar()    # he hates indents smaller than 4

[Martijn Faassen]
> Agreed. The Zope sources are full of the former and that idiom
> doesn't help me read them particularly.
>
> Perhaps 1.6 could introduce this requirement, and some scripts to
> convert existing code. The only use I can see for the former is to
> write nice obfuscated one liners. :)

There are a few things in the language primarily for convenience in
interactive mode.  This is one of them.  Being Dutch, Guido grossly
overestimated the common sense of his American audience <wink>.

my-people-invented-bad-taste-and-proud-of-it-ly y'rs  - tim





From haskin at ptway.com  Mon Nov  8 16:52:38 1999
From: haskin at ptway.com (Brian Haskin)
Date: Mon, 08 Nov 1999 16:52:38 -0500
Subject: Stackless python progress?
Message-ID: <38274626.5D8C1EC2@ptway.com>

Just wondering if the work on stackless python is still progressing and
if  it looks like it will be integrated into the regular distribution in
the foreseeable future? also has there been any work on combining
'microthreading' and the stackless stuff?

Brian Haskin



From wtanksle at hawking.armored.net  Thu Nov  4 16:01:58 1999
From: wtanksle at hawking.armored.net (William Tanksley)
Date: 4 Nov 1999 21:01:58 GMT
Subject: Some more Ruby observations (Re: Comparison between Python and "Ruby")
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp> <7vjtrc$hmm$1@newshost.accu.uu.nl>  <3821A860.5D32AFCC@compaq.com> <slrn824eo2.lsi.neelk@brick.cswv.com>
Message-ID: <slrn824lv6.ngk.wtanksle@hawking.armored.net>

On 5 Nov 1999 01:22:23 GMT, Neel Krishnaswami wrote:
>Greg Ewing <greg.ewing at compaq.com> wrote:
>>Here are some random observations from studying the Ruby
>>source code:

>>* It doesn't use bytecode. The evaluator seems to operate
>>  directly on a parse tree. I don't know what implications
>>  this has for execution speed.

>I was under the impression that Ruby did tail-recursion elimination,
>and since continuation-passing forms are just a hair away from
>bytecode, I am a bit surprised by this news. I'm not saying this is
>impossible (witness the scm Scheme interpreter), mind, just a little
>surprised.

>Are there any syntax-tree munging phases to do stuff like common
>subexpression elimination and loop hoisting? I guess that could yield
>enough speedups to make a VM not yet necessary.

Parse trees tend to evaluate faster than bytecode when both have only
primitive optimizations.  They are also much, much easier to optimize.  In
general, I would expect a parse tree implementation to outperform a
bytecode implementation given equivalent work.

>Neel

-- 
-William "Billy" Tanksley



From Gareth.McCaughan at pobox.com  Wed Nov 24 17:20:56 1999
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: 24 Nov 1999 22:20:56 +0000
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <867lj85g4c.fsf@g.local>
	<al8yabo9fm3.fsf@sirppi.helsinki.fi>
Message-ID: <86so1v4jlj.fsf@g.local>

Markus Stenberg wrote:

[I said:]
>>   - Some more iteration constructs would be nice.
>> 
>>       - do ... while
> 
> ugh. matter of taste I suppose, but I don't find that lack big deal.

The main reason I'd like to see it there is that I use
Python sometimes for teaching to (young -- 12-15-ish)
novices, and "while 1: ... break ..." is a little hard
to motivate.

>>       - parallel iteration across more than one list
> 
> for list1item, list2item, list3item in map(None, list1, list2, list3):
> 	print 'whee', list1item, list2item, list3item

Yeah, I know. But if I want that sort of thing I can use C. :-)

>>       - a less yucky way of saying "for i in range(1,20,3)"
> 
> see above - although some special syntactic sugar would be "nice", I prefer
> minimalistic approach over say, C.. of course, Dylan does it (IMNSHO)
> cleanest:
> 
> for (i from 3 to 19 by 3)
> 	...
> 	
> 
> Regrettably I can't tolerate Dylan's whitespace approach :P
> (I write accidentally too often stuff like i-j, as opposed to i - j)

Dylan is sort of nice. I can't think of any purpose for which
I'd prefer Dylan to Common Lisp, though; the strengths of the
two languages are in similar places, and I think CL wins.

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
sig under construction



From bjorn at roguewave.com  Mon Nov 29 12:00:59 1999
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Mon, 29 Nov 1999 10:00:59 -0700
Subject: Python bugs
Message-ID: <51C9F3C392D0D111A15600A0C99E64A9AF5690@localhost>

Just found two bugs:

	- os.environ and shelve does not conform to the 
	  Mapping Type interface (2.1.6)

(and in case you were wondering, I'm not interested in workarounds, but
rather not having to special case my code <wink>)

--bjorn

Python 1.5.2 (#0, Apr 13 1999, 10:51:12) [MSC 32 bit (Intel)] on win32
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import os
>>> {}.update(os.environ)
Traceback (innermost last):
  File "<stdin>", line 1, in ?
TypeError: dictionary, instance
>>> import shelve
>>> s = shelve.open('foo')
>>> s.get('bar', 'baz')
Traceback (innermost last):
  File "<stdin>", line 1, in ?
AttributeError: get
>>>





From friedrich at pythonpros.com  Tue Nov  2 13:20:12 1999
From: friedrich at pythonpros.com (Robin Friedrich)
Date: Tue, 02 Nov 1999 18:20:12 GMT
Subject: Template for CGI
References: <7vn0sk$7cp$1@news.fsu.edu>
Message-ID: <7vna0o$9rt$1@nnrp1.deja.com>

That is a feature of TemplateDocument() in the HTMLgen package. You
might try downloading it from http://starship.python.net/lib.html




In article <7vn0sk$7cp$1 at news.fsu.edu>,
  "Glenn Kidd" <gek8365 at garnet.acns.fsu.edu> wrote:
> I am a newbie to Python and I have a question.  I want a CGI script
that
> will read a file template and replace certain words with variables.
Ii
> tried putting the variable names in the file but it is read as palin
text.
> Is there a way to embed variables into a teplate or a replace
function in
> Python?  Thank you for any help you give.
>
> Glenn Kidd
>
>


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



From thomas.heller at ion-tof.com  Mon Nov 29 13:24:37 1999
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Mon, 29 Nov 1999 19:24:37 +0100
Subject: Need help with html files
References: <3842bd0e.23017938@122.0.0.250>
Message-ID: <024601bf3a96$fe604290$4500a8c0@thomasnotebook>

> hi,
>
> i would like to open a local html file from within python,
> When the html is opened, it should be opened like if you double click
> on the filename, so it uses the know filetype from windows.
>
> in short , when someone has netscape as the default browser.
> I open the html file python should call the html file , and windows
> opens the default browser with the file as argument.
>
> I hope this is clear enough, at the moment i can?t think of a way to
> explain it differently.
>
> if you know what i mean and if you have a solution for my problem
> please post or mail.
>
> thanks
>
import os
os.system ("start c:\\myfile.html")

Thomas Heller




From dgrisby at uk.research.att.com  Wed Nov 10 07:40:57 1999
From: dgrisby at uk.research.att.com (Duncan Grisby)
Date: 10 Nov 1999 12:40:57 -0000
Subject: omniORB for Python pre-release now available
Message-ID: <80bp4p$fn$1@pineapple.uk.research.att.com>

AT&T Laboratories Cambridge is pleased to announce the availability of
the first public pre-release of omniORB for Python. omniORB is a free
high-performance CORBA ORB.

The bindings adhere to the latest draft Python mapping available from
the OMG. Although this is a pre-release, the ORB has no known bugs,
and is in daily use in our lab.

omniORBpy is free software. The ORB runtime is released under the
terms of the GNU Lesser General Public License; the associated tools
are under the GNU General Public License.

omniORBpy is a wrapper around the omniORB core library, and is
therefore reasonably fast. Although most of the ORB is written in C++,
the stubs generated from IDL are pure Python.

This release of omniORBpy is also the first outing for omniORB's new
IDL compiler, which has a simple Python back-end interface. If you
have ever entertained the idea of generating some special-purpose code
from your IDL, but been scared off by the complexity of available IDL
front-ends, you may want to look at the new compiler.

For more information, and to download omniORBpy, visit

  http://www.uk.research.att.com/omniORB/omniORBpy/

-- 
 -- Duncan Grisby  \  Research Engineer  --
  -- AT&T Laboratories Cambridge          --
   -- http://www.uk.research.att.com/~dpg1 --



From mhammond at skippinet.com.au  Mon Nov 29 23:57:53 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Tue, 30 Nov 1999 04:57:53 GMT
Subject: Access violation while trying to replace stdout with a local disk file
References: <1268224356-30441862@hypernet.com> <000a01bf3a98$4f13c940$36f96d8c@sylvia>
Message-ID: <lPI04.9650$VA6.47875@news-server.bigpond.net.au>

SINICA wrote in message <000a01bf3a98$4f13c940$36f96d8c at sylvia>...
>
>> They both need to be linked to the same c-runtime library, (ie,
>> "Multithreaded DLL" in Settings | Compile | Code Generation).
>>
>> - Gordon
>>
>>
>
>i've changed the c-runtime to "Multithreaded " for myprog.exe
>and "Multithreaded DLL" for python15.dll
>but it still refuse to work,
>shall i change to "Multithreaded DLL" instead of "Multithreaded" for
>myprog.exe ?

Yes - as Gordon said - "They both need to be linked to the same c-runtime
library, (ie,
"Multithreaded DLL" in Settings | Compile | Code Generation)."

Mark.





From 0498730983 at 03840938.net  Mon Nov  1 01:44:33 1999
From: 0498730983 at 03840938.net (0498730983 at 03840938.net)
Date: Mon, 01 Nov 1999 06:44:33 GMT
Subject: HP Inkjet Blowout
Message-ID: <BRQ79UK3.430KT4GO@03840938.net>

				The Inkjet Company has the following inkjet specials:
				All Prices Include Free Shipping and Free Sales Tax





EPSON

Epson -  $8.00 - Black  or 10 for $65.00 - All Models*
Epson -  $12.00 - Color or 10 for $100.00 - All Models*
*Epson - $16.00 - Black or 10 for $135.00 - 900/3000
*Epson - $19.00 - Color or 10 for $165.00 - 750/900/1200/3000


CANON

Canon BCI-21B - $5.00 - Black BJC-2500/4000/4400/5000
Canon BCI-21C - $8.00 - Color BJC-2500/4000/4400/5000
Canon BC-20 - $18.00 - Black BJC-2500/4000/4300/4400/
Canon BC-02 - $16.50 - Black BJC-210/230/BJ-200
Canon BC-05 - $28.00 - Tri-Color BJC-210/240
Canon BX-3 - $28.00 - Black B-540/550/640

(See Below For Apple & H.P. Products)


				
				The Inkjet Company

				1-310-517-9999
				1-503-905-6652 Fax


APPLE

M8041G/C - $17.00 - Black Styler Writer 1500
M4609G/C - $30.00 - Tri-Color Styler Writer 1500
M3240G/A - $19.00 - Large "Double" Black Style Writer 2400
M3330G/A - $5.00 - Black Style Writer 2400
M3329G/A - $9.00 - Tri-Color Style Writer 2400 
M1949G/A - $3.75 - Cyan Color Style Writer Pro 
M1950G/A - $3.75 - Magenta Color Style Writer Pro
M1951G/A - $3.75 - Yellow Color Style Writer Pro
M1952G/A - $3.75 - Black High Capcity Color Style Writer Pro


HEWLETT PACKARD

51625A - $19.00 - Tri-Color Deskjet 500c/550 Series
51626A - $16.50 - Black Deskjet Series / OfficeJet 300 / HP Fax
51629A - $16.50 - Black Deskjet 600 Series / OfficeJet 500 Series
51649A - $19.00 - Tri-Color Deskjet 600 Series / OfficeJet 500 Series
51645A - $16.50 - Black Deskjet 700/800/1000/1600
51640A - $16.50 - Black Deskjet 1200 / CopyJet



					The Inkjet Company

					1-310-517-9999
					1-503-905-6652 Fax

				    Visa, Mastercard, Amex, Discover
				Fast Free Delivery - usually in two days or less!



From joecot1 at bellatlantic.net  Sat Nov 13 22:54:49 1999
From: joecot1 at bellatlantic.net (Joe Cotellese)
Date: Sun, 14 Nov 1999 03:54:49 GMT
Subject: Python and sending emails with it
References: <382E2DB2.A00B9109@nv2.cc.va.us>
Message-ID: <942551361.1861452872@news.bellatlantic.net>

Well it depends.  You can generally use the SMTPLIB module for sending mail. 
It works pretty well.  I have however found it to have problems sending mail to
a Worldmail (Eudora) server.

The library is smtplib.py and it has a usage sample at the bottom of the module.

Regards,
Joe Cotellese

 Jeremy Misavage <tekhir at yahoo.com> wrote: > I'm kinda new to
python and I'm trying to write a small program that will take > in form data,
save it in a db, and then send an email to a person. >  > The program processes
a users request for a "equipment request."  It gets > loggged in a db (already
done) and sends an email to the person in change of the > equipment.  I need
help on the email part.  Does some out there have some code > that will taken
data and send it to a person.  It shouldn't need a username and > password, but
if it does I'll live with that.




From andy at robanal.demon.co.uk  Sun Nov  7 11:42:59 1999
From: andy at robanal.demon.co.uk (Andy Robinson)
Date: Sun, 07 Nov 1999 16:42:59 GMT
Subject: Drawing (directed) graphs
References: <000a01bf2792$b83ff150$3acbd9c2@peridot.optichrome.com> <y0jeme5owc9.fsf@vier.idi.ntnu.no> <14370.59564.979979.285996@gargle.gargle.HOWL> <y0jbt99ovuw.fsf@vier.idi.ntnu.no>
Message-ID: <3826abe1.17176077@news.demon.co.uk>

mlh at vier.idi.ntnu.no (Magnus L. Hetland) wrote:

>Charles G Waldman <cgw at fnal.gov> writes:
>
>> Magnus L. Hetland writes:
>[...]
>>  > Even on Minix you could easily generate jpegs, or postscript files or
>>  > whatever with Piddle. If you can't run an X server or something, that
>>  > only means that you can't see it... Or maybe you have a printer?
>> 
>> Or for maximum portatbility, you could use the "aa" ascii-art library
>> to display the images on an ASCII terminal.... it's a C library, but
>> it would be trivial to SWIG up a set of Python bindings, if this
>> hasn't been done yet.
>
>Excellennt suggestion :)
>
>A nice backend for PIDDLE... Any takers?

No way - Joe would hold up the release forever until we managed
rotated text!

- Andy



From evan at tokenexchange.com  Thu Nov 11 13:14:08 1999
From: evan at tokenexchange.com (Evan Simpson)
Date: Thu, 11 Nov 1999 12:14:08 -0600
Subject: What are Aspects? (Re: How can I make my assertions smarter?)
References: <Pine.WNT.4.05.9911032042440.109-100000@david.ski.org> <7vvq63$ehm$1@nntp4.atl.mindspring.net> <3826C32A.FB47EF6@compaq.com> <slrn82h5o6.26j.wtanksle@hawking.armored.net> <80eocf$8rc$1@nntp5.atl.mindspring.net>
Message-ID: <WIDW3.331$Ae3.219038@news-dal.corridex.com>

<disclaimer>All I know about Aspects I learning in a half-hour of
browsing</disclaimer>

I think that this depends on the language in which you write your Aspects (I
don't mean C/Python/etc, I mean the special Aspect-specification language).
In the examples I looked at, pattern-matching on the class, method, and
method arguments determined where each part of an Aspect would be applied.
Effectively, you could say thing like:

o Insert this code before the start of all methods of every class
o Put this code in a try...finally wrapped around the body of any method
named "foo" in any class
o Insert this code after the end of all constructors of classes "Fee",
"Fie", and "Foe"

It all looks very interesting and powerful, but I haven't the faintest idea
how it could be applied to a dynamic language like Python.  From what I
could tell, the existing implementation depends on all classes and methods
being defined and fixed at compile time.

Aahz Maruch <aahz at netcom.com> wrote in message
news:80eocf$8rc$1 at nntp5.atl.mindspring.net...
> I feel like I'm still missing something.  What if you only want half of
> your classes to get ref-counting?  Or are you saying that aspects are by
> definition applied across *all* code, and that they're more like hooks
> into the Python interpreter?

Cheers,

Evan @ 4-am





From fredrik at pythonware.com  Mon Nov 29 06:30:51 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 29 Nov 1999 12:30:51 +0100
Subject: help with opening html files
References: <384261c4.8540090@122.0.0.250>
Message-ID: <028601bf3a5d$31abdea0$f29b12c2@secret.pythonware.com>

volumeregeling <volumeregeling at hotmail.com> wrote:
> Can anyone tell me how to open a html file from within python in
> windows.

html files are text files, and can be handled
in the usual way.  see chapter 7 in the tutorial
for more info:

http://www.python.org/doc/current/tut/tut.html

</F>




From quinn at groat.ugcs.caltech.edu  Fri Nov 26 17:54:27 1999
From: quinn at groat.ugcs.caltech.edu (Quinn Dunkan)
Date: 26 Nov 1999 22:54:27 GMT
Subject: python test frameworks
Message-ID: <slrn83u3um.4ge.quinn@groat.ugcs.caltech.edu>

I know this sort of thing has been talked about before, but I'm looking for
some information on techniques for self-testing pyhon code (something I
should have been doing before, but only now am I writing a system complex
enough to warrant it).  I remember Tim's thread about his doc-string embedded
testing, and that looked interesting, but I can't find it now.  And links to
discussions of the merits of different strategies would be good, too.

thanks



From m.faassen at vet.uu.nl  Thu Nov 18 16:00:46 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 18 Nov 1999 21:00:46 GMT
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <JlFX3.3659$s87.67930@dfiatx1-snr1.gtei.net> <80oqn9$ocq$2@newshost.accu.uu.nl> <oLUX3.286$yA4.4821@dfiatx1-snr1.gtei.net> <80p9su$ppo$2@newshost.accu.uu.nl> <80piru$13rl$1@hub.org> <80re5d$55a$1@newshost.accu.uu.nl> <fjPY3.770$An3.14793@newsfeed.avtel.net> <810jto$lo0$1@newshost.accu.uu.nl>
Message-ID: <811pdu$dvo$1@newshost.accu.uu.nl>

Martijn Faassen <m.faassen at vet.uu.nl> wrote:
[snip]
> No, no, a more extensive deja search indicated to me that this person has
> posted under other names to the group previously. At least, unless there are
> 10 people with the same speech patterns posting to comp.lang.python. :)

with the same speech patterns _from vex.net_, I meant! deja nor I are that
good at scanning for similar speech patterns yet, but the vex.net subset
is doable for the eyeball. 

Still-curious-ly yours,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?



From mlh at vier.idi.ntnu.no  Tue Nov  9 08:38:56 1999
From: mlh at vier.idi.ntnu.no (Magnus L. Hetland)
Date: 09 Nov 1999 14:38:56 +0100
Subject: pythonistas or pythoneers?
References: <07111999.2@sanctum.jae> <382633B1.50993D93@callware.com> <y0jn1spni7c.fsf@vier.idi.ntnu.no> <806unr$6qp$1@nnrp1.deja.com>
Message-ID: <y0jln87239r.fsf@vier.idi.ntnu.no>

init__self at my-deja.com writes:

> > No way - we are all Pythoneers at the frontiers for computer
> > science... ;)
> 
> Wouldn't that be a Pyoneer then?
> 

<chuckle>

Guess so... But the overlap with the string "Python" is only two
characters long...  (But the editing distanc between Python and Pyon
is only 2...)

--

  Magnus          Echelon jamming noise:
  Lie             FBI CIA NSA Handgun Assault Bomb Drug Terrorism
  Hetland         Special Forces Delta Force AK47 Hillary Clinton 



From andrew.markebo at telelogic.se  Wed Nov  3 03:09:19 1999
From: andrew.markebo at telelogic.se (Andrew Markebo)
Date: 03 Nov 1999 09:09:19 +0100
Subject: dejasearch.py - an www.deja.com client, Alpha version comments wanted
Message-ID: <ksk8o06low.fsf@telelogic.se>

Well my dejasearch.py seems to be usable now.. readme and link below

        /Andy

Can be found at:
        http://www.fukt.hk-r.se/~flognat/DejaSearcher2.zip

dejasearch.py - client for deja.com

Allows you to easily see all the headers that matches your search,
downlad/view the articles you are interested in, and save your query
so you can load it again, and continue downloading articles.

This is still some kind of alpha-version would like your comments
about the GUI, if you have any ideas what to use instead and so on.

In the article header list you see a 'status' letter first, it means
A:Away, not downloaded, D:Downloading, H:Home, you can see the article
if you doubleclick it.

What works:
* Search on keywords
* download selected
* load/save
* doubleclick on article header line to download article, if
  downloaded it views the article in the article frame.. or ;-)

Todo:
* Use asyncore.py instead of BGFetcher (stability)
* implement the selection fields
* look why I only can do 'default' searches on keywords.
* Improve the gui
* Incorporate your changes ;-)

      /andy
	flognat at fukt.hk-r.se




From smigowsky at dspace.de  Tue Nov 30 11:17:06 1999
From: smigowsky at dspace.de (Stefan Migowsky)
Date: Tue, 30 Nov 1999 17:17:06 +0100
Subject: COM and Powerpoint control
Message-ID: <326536345498D311B3BC00105A39802A074500@newsintern.dspace.de>

Hi, 
 
just use SlideShowWindows.Next, e.g in your code :

from win32com.client import constants, Dispatch

x = Dispatch('Powerpoint.Application.8')
x.Presentations.Open(FileName='c:\temp\COMTutorial.ppt')
x.ActivePresentation.SlideShowSettings.Run()

print " Show first slide for 5 secs"
win32api.Sleep(5000)

print " Show next slide for 5 secs"
x.SlideShowWindows(1).View.Next()

print " Show next slide for 5 secs"
win32api.Sleep(5000)
.
.
.
print " That's all folks"
x.SlideShowWindows(1).View.Exit()


Stefan
--------------------------------------------------------
mailto:smigowsky at dspace.de 		http://www.dspace.de

>-----Original Message-----
>From: Harald Singer [mailto:singer at itl.atr.co.jp]
>Sent: Tuesday, November 30, 1999 3:09 PM
>To: python-list at python.org
>Subject: COM and Powerpoint control
>
>
>Hi,
>
>I am looking for an example of an "outside controlled" PowerPoint
>presentation using COM. My final goal is to time-align the slides
>presentation with an audio file, because the PowerPoint internal
>slide-show timing is massively broken.
>
>The following code snippet worked promisingly well on Win98:
>
> from win32com.client import constants, Dispatch
>
> x = Dispatch('Powerpoint.Application.8')
> x.Presentations.Open(FileName='c:\temp\COMTutorial.ppt')
> x.ActiveWindow.ViewType = constants.ppViewNotesPage
>
>Now, I tried to record a macro to see what commands might be used in
>slideshow mode, but looking at the generated VB code, I can't see any
>commands that advance the slideshow.
>
>-- VB snippet start ---
>Sub Macro1()
>'
> ppShowAll
>        .AdvanceMode = ppSlideShowUseSlideTimings
>        .PointerColor.SchemeColor = ppForeground
>        .Run
>    End With
>    SlideShowWindows(Index:=1).View.Exit
>End Sub'
>-- VB snippet end ---
>
>Any help or code fragments are appreciated. BTW, I did a search
>for this but nothing relevant came up.
>-- 
>Harald Singer	ATR-ITL		singer at itl.atr.co.jp	++81-774-95-1389
>
>
>-- 
>http://www.python.org/mailman/listinfo/python-list
>



From cjensen at be-research.ucsd.edu  Wed Nov 17 13:23:53 1999
From: cjensen at be-research.ucsd.edu (Curtis Jensen)
Date: Wed, 17 Nov 1999 10:23:53 -0800
Subject: Multi-Dimensional Arrays.
Message-ID: <3832F2B9.F64C8AD7@be-research.ucsd.edu>

I'm working on a Python user interface that will be connected to some
fortran code with heaps of multi-dimensional arrays.  I can't seem to
find any references to multi-dimensional arrays in Python.  Is there an
easy way to create multi-dimensional arrays in Python?

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


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/19991117/396e368f/attachment.html>

From jam at quark.emich.edu  Wed Nov 10 18:49:26 1999
From: jam at quark.emich.edu (Jeff)
Date: Wed, 10 Nov 1999 18:49:26 -0500
Subject: Intelligent strptime?
In-Reply-To: <19991110204023.A8220@optiplex.palga.uucp>
References: <19991110171256.A5196@optiplex.palga.uucp> <14377.41778.712189.7874@gargle.gargle.HOWL> <19991110204023.A8220@optiplex.palga.uucp>
Message-ID: <19991110184926.B8734@quark.emich.edu>

On Wed, Nov 10, 1999 at 08:40:23PM +0100, Gerrit Holl wrote:
[..snipped discussion about how best to handle the 'various date format'
problem..]
> 
> It could return a tuple or something then. But if I want people to choose:
> either input "Nov, 23" or "12/6/99" or "1 December 2000", I don't want to
> check too much. There are much date processing programs (date, I think) that
> do this intelligent. You could also let the above think depend on the timezone,
> if it's an american timezone, take the american way, else, take the
> european way. But returning a tuple on 132 days seems better to me.
> 

I ported a small piece of code from the CVS sources into a python module
that returns the number of seconds since the Epoch instead of a tuple. I
called it 'getdate', and released it a while ago to my startship.python.net
account (ftp://starship.python.net/pub/crew/jam/). 

the module is comprised of one function that is able to act intelligently
when given various date formats like the ones you are describing and seems
to make the 'right' choice 'most' of the time. please try it out-- it might
suit your purposes nicely, especially if you are writing an interactive
application that accepts date input from a human user.

here is an example of this module in action:

[~] [6:36pm] [jam at toast-pts/3] % python
Python 1.5.2 (#1, Apr 18 1999, 16:03:16)  [GCC pgcc-2.91.60 19981201
(egcs-1.1.1  on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import getdate
>>> dir(getdate)
['_RCSID', '__doc__', '__file__', '__name__', 'error', 'getdate', 'parse']
>>> getdate.getdate("1 december 2000")
975646800L
>>> getdate.getdate("1999-12-01")
944024400L
>>> getdate.getdate("12/1/1999")
944024400L
>>> getdate.getdate("last friday 15:00 PST")
941842800L
>>> getdate.getdate("now")
942277099L
>>> import time
>>> time.strftime("%Y-%m-%d %H:%M %Z", time.localtime(getdate.getdate("now")))
'1999-11-10 18:39 EST'
>>> time.strftime("%Y-%m-%d %H:%M %Z", time.localtime(getdate.getdate("yesterday")))
'1999-11-09 18:39 EST'
>>> time.strftime("%Y-%m-%d %H:%M %Z", time.localtime(getdate.getdate("last friday 15:00 PST")))
'1999-11-05 18:00 EST'
>>> time.strftime("%Y-%m-%d %H:%M %Z", time.localtime(getdate.getdate("8pm")))
'1999-11-10 20:00 EST'

please let me know if there are any questions or concerns.

regards,
J
-- 
|| visit gfd <http://quark.emich.edu/>
|| psa member #293 <http://www.python.org/> 
|| New Image Systems & Services, Inc. <http://www.newimage.com/>



From krutz at cg.tuwien.ac.at  Wed Nov 24 13:59:29 1999
From: krutz at cg.tuwien.ac.at (Markus Krutz)
Date: Wed, 24 Nov 1999 18:59:29 GMT
Subject: newbie question
References: <943469319.838266356@news.teleweb.at>
Message-ID: <943469465.1530627167@news.teleweb.at>

On Wed, 24 Nov 1999, Markus Krutz wrote:
>How the f*** do I create a string containing a *single* "\" ?
>
>regards
>Markus

"Problem" solved. Damned interpreter output ;)



From Gaetan_Corneau at baan.com  Fri Nov 26 09:03:04 1999
From: Gaetan_Corneau at baan.com (Gaetan Corneau)
Date: Fri, 26 Nov 1999 09:03:04 -0500
Subject: The First Annual Obfuscated Python Contest
Message-ID: <816010E2456BD111A48700805FBBE2EEFDF83B@ex-quebec-u1.baan.com>

Anyone has ever seen an elegance Python contest?
I mean, tough problems cracked with simple, elegant code? :)

______________________________________________________
   Gaetan Corneau
   Software Developer (Quality Assurance Team)
   BaaN  Supply Chain Solutions  
   http://www.baan.com
   E-mail: Gaetan_Corneau at baan.com                  
   Tel: (418) 266-8252          
______________________________________________________
"Profanity is the one language all programmers know best"





From gvwilson at nevex.com  Sun Nov 28 13:51:38 1999
From: gvwilson at nevex.com (Greg Wilson)
Date: Sun, 28 Nov 1999 13:51:38 -0500
Subject: pre/post-condition support?
Message-ID: <_Re04.8705$cz3.315483@nnrp1.uunet.ca>

Has anyone experimented with Python extensions to support programming by
contract (pre/post conditions on methods + state assertions that are
respected by inheritance)?  If so, I'd welcome pointers.

Thanks,

Greg

p.s. I'm aware this can be done by hand by putting assertions at the start
and end of methods.  However, that doesn't automatically work right when
methods have multiple exit points (return statements, exceptions that
callers are expected to catch, etc.), and isn't respected by inheritance.






From phd at sun.med.ru  Wed Nov 10 08:35:54 1999
From: phd at sun.med.ru (Oleg Broytmann)
Date: Wed, 10 Nov 1999 16:35:54 +0300 (MSK)
Subject: time/strptime/mktime/localtime
Message-ID: <Pine.GSO.4.20.9911101632390.1674-100000@sun>

Hello!

   The following program prints "1999-09-21 21:44:33" on Linux and
"1999-09-21 22:44:33" on Solaris (1 hour offset).

   Is it a problem with my config? libs? python? Timezone is Europe/Moscow,
set (I think) correctly on both systems.

#! /usr/local/bin/python -O

import time

t = time.strptime("1999-09-21 21:44:33", "%Y-%m-%d %T")
m = time.mktime(t)
print time.strftime("%Y-%m-%d %T", time.localtime(m))

Oleg.
---- 
    Oleg Broytmann  National Research Surgery Centre  http://sun.med.ru/~phd/
           Programmers don't die, they just GOSUB without RETURN.




From Gareth.McCaughan at pobox.com  Thu Nov 25 15:51:14 1999
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: 25 Nov 1999 20:51:14 +0000
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <867lj85g4c.fsf@g.local>
	<al8yabo9fm3.fsf@sirppi.helsinki.fi> <86so1v4jlj.fsf@g.local>
	<al8hfiaan8p.fsf@sirppi.helsinki.fi>
Message-ID: <86ln7m2t31.fsf@g.local>

Markus Stenberg wrote:

> I have always found lisp-ish syntax of "masses of ()s" disconcerting in
> large amounts - admittedly, I've written more than few lines of
> [ce]lisp/scheme/.., but never enjoyed browsing the results very much. I
> consider Dylan to be "readable" clisp.

One quickly gets used to Lisp syntax. Like using indentation to
denote block structure. :-) After a while, you just stop noticing
all those parentheses.


> From trivial example - do you consider (+ 2 (* 3 4)) more readable than 2 +
> 3 * 4 most languages (including Dylan) allow me to do?

No, I don't. But I don't consider it hugely less readable,
and that really is the *worst* case for Lisp syntax. Really
complicated expressions are about equally readable either
way; and Lisp's great (excessive?) simplicity makes possible
its great macro system. And *that* lets me make a lot of
things much more readable than they otherwise would be.

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
sig under construction



From thantos at chancel.org  Thu Nov 18 18:39:03 1999
From: thantos at chancel.org (Alexander Williams)
Date: Thu, 18 Nov 1999 23:39:03 GMT
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <80pe77$osc$1@hub.org> <slrn830he9.n28.mgm@unpkhswm04.bscc.bls.com> <80vp2m$275$1@nntp9.atl.mindspring.net>  <slrn838ohs.fv8.mgm@unpkhswm04.bscc.bls.com>
Message-ID: <slrn8393kv.f1h.thantos@chancel.org>

On 18 Nov 1999 20:31:58 GMT, Mitchell Morris
<mgm at unpkhswm04.bscc.bls.com> wrote:

>Reading is not so bad ... searching blows. As an example, I'm looking at the
>page for the "re" module and want to see the argument list for "unpack" but
>can't seem to remember if it's in "string" or "struct" or
>Guido-only-knows-where. Lynx provides squat for this operation, so I wind up

Magic word: index.

See the last page of the doc list in HTML, an index of functions and
terms hyperlinked to the discussion thereof.  Voila.

Has the dimmunation of print media really caused us to forget such
simple methods?

-- 
Alexander Williams (thantos at gw.total-web.net)
"In the end ... Oblivion Always Wins."



From gerrit.holl at pobox.com  Fri Nov 19 07:58:11 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Fri, 19 Nov 1999 13:58:11 +0100
Subject: Python in business programming [was Re: Beginning programming with Python]
In-Reply-To: <1269336472-10033930@hypernet.com>
References: <383195C1.826B3342@iqsoft.hu> <1269336472-10033930@hypernet.com>
Message-ID: <19991119135811.B4306@optiplex.palga.uucp>

Gordon McMillan wrote:
> (the code, compile, test cycle is now rarely more than half an hour).

Hehe, Win NT should be rewritten in Python; it seems that this cycle costs
30 hours for NT :-))

> Test suites are 
> more important (but also easier). For large systems, 
> decomposing the problem cleanly is very important - to 
> preserve your sanity you need the smallest, cleanest 
> interfaces possible between subsystems. But, if at first you 
> foul it up, you can still rewrite most of it before lunch.

I agree. But some functions are too small to be used:
in os.path:
def os.path.isabs(s):
    return p[:1] == '/'

I think that's a redicolous (how does one spell that?) function...

regards,
Gerrit.

-- 
"We all know Linux is great...it does infinite loops in 5 seconds."
(Linus Torvalds about the superiority of Linux on the Amterdam
Linux Symposium)



From rp2885 at email.sps.mot.com  Thu Nov 11 17:04:41 1999
From: rp2885 at email.sps.mot.com (Aaron Ginn (rp2885))
Date: Thu, 11 Nov 1999 15:04:41 -0700
Subject: Beginning programming with Python
References: <sng0ychgwt.fsf@motorola.com> <382B3A53.5816B2@callware.com>
Message-ID: <382B3D79.74CD3E0F@email.sps.mot.com>

Ivan Van Laningham wrote:

> Hi All--
>
> Aaron Ginn wrote:
> >
> > Hello all,
> >
> > I'm starting a non-profit organization dedicated to promoting computer
> > literacy in low-income youth (ages 12-18).
>
> An admirable goal indeed.  I will be interested in hearing how well this
> works.  Are you affiliating yourself with any municipal, civic or
> religious organizations?

I've created a website for our organization at
http://www.mindspring.com/~alginn/gwam.html

The site is a little out of date at this point.  The original goal was to use
only Linux as the base OS for this project.  However, one of the people I've
recruited to work on this is a C.S. professor at Arizona State who has a contact
in the Microsoft Education office.  We hope that Microsoft will supply us with
free licenses so that we can put dual-boot machines together for the kids.  If
they do, I'll definately have to change our logo! :)

All of the people I've recruited in this go to my church, and we will use the
church as our meeting place, but we have no religious agenda.


>
>
> > The curriculum will
> > include an introduction to programming using either Python or Perl.
> >
> > First, I know Perl very well, but I know nothing about Python.  I've
> > heard from many sources that Python is the best language to teach
> > someone with no programming experience because it emphasizes
> > fundamentals and the correct (i.e. best) way to program.  Could anyone
> > comment on how the two languages compare in this aspect?  Also, I will
> > need to learn Python as well if I choose to go that route.  I am very
> > literate with Perl and Tcl/TK.  How quickly can I expect to pick up
> > Python?
> >
>
> *Extremely* quickly.  In fact, if you'd started yesterday you wouldn't
> need to ask this question.

That's what I expected.  I perused the books in question and musc of the syntax
looked _exactly_ like Perl.

An honest opinion, please; is Python better suited for beginners than Perl, as
I've heard?


>
>
> > Secondly, if I choose to use Python for this, I would like to supply
> > each student with a reference book on Python.  I've narrowed it down
> > to a choice between the two O'Reilly books, "Programming Python" and
> > "Learning Python".  PP apears to have a lot more meat to it, but LP
> > appears to be more suited to the beginner.  Assuming the audience
> > includes kids that have never programmed before, should I go with LP,
> > or will I miss out on anything by not getting the more detailed book?
> >
>
> Of the two books mentioned, yes, _Learning Python_ would be the better
> choice.  However, if you can wait, _Teach Yourself Python in 24 Hours_
> is aimed *specifically* at people who have never programmed before.  It
> covers:
>
>         1)  Introduction to programming (*rock-bottom* fundamentals--what's a
> variable? What's a for loop?)
>         2)  Introduction to Object-oriented programming
>         3)  Introduction to Tkinter

>
>
> The three parts have 8 chapters each.
>
> TYPython will be out early next year.  When is your course starting?
> Where is Chandler?-)

We are planning on starting our training in March, so I can wait a while.  I'll
probably pick up a copy of PP for myself in the meantime.

Chandler is a suburb of Phoenix.  Actually, I live 1 mile from the Phoenix
border.


>
>
> PP is very thorough, but needs a new edition.
> LP is aimed at people who do have some idea of how to program, and is a
> gentle way to further your programming skills.
> Both books are funny and interesting, too.
>
> TYPython is for total newbies; I can't speak to either funny or
> interesting (modesty overwhelms me), but I try to get their attention.
>
> None of these three books is a *reference*, however.  If you really want
> a reference, then you're better off with Dave Beazley's _Python:
> Essential Reference_.  But you need to know Python before you can make
> use of it.

Reference was a bad choice of words.  I want a book that will introduce the kids
to basic programming fundamentals and will provide examples of how to do things
in a given language that is easy to learn.  I don't want "Python in a Nutshell"
if such a thing even exists.


>
>
> Hope this helps,

It does.  Thanks a lot!

Aaron

>
>
> <oh-goodness-i-am-so-shy-i-embarass-myself>-ly y'rs,
> Ivan
>
> <or-is-that-shameless-shameless-plugs-r-us?>-ly y'rs,
> Ivan
> ----------------------------------------------
> Ivan Van Laningham
> Callware Technologies, Inc.
> ivanlan at callware.com
> ivanlan at home.com
> http://www.pauahtun.org
> See also:
> http://www.foretec.com/python/workshops/1998-11/proceedings.html
> Army Signal Corps:  Cu Chi, Class of '70
> Author:  Teach Yourself Python in 24 Hours
> ----------------------------------------------



--
Aaron J. Ginn                     Motorola SPS
Phone: (480) 814-4463             SemiCustom Solutions
Fax:   (480) 814-4058             1300 N. Alma School Rd.
mailto:aaron.ginn at motorola.com    Chandler, AZ 85226






From neelk at brick.cswv.com  Tue Nov  2 22:42:21 1999
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 3 Nov 1999 03:42:21 GMT
Subject: Comparison between Python and "Ruby"
References: <1270663667-24391282@hypernet.com> <381ED7A6.D7D7FB13@compaq.com> <877lk1wiij.fsf@ev.netlab.co.jp> <19991102154543.A4469@optiplex.palga.uucp>  <381F9EB4.63FA033F@python.net>
Message-ID: <slrn81ve6m.l1c.neelk@brick.cswv.com>

Thomas A. Bryan <tbryan at python.net> wrote:
>
>Yes, there are.  Larry Wall's speeches (Wall is the Guido of Perl)
>often proclaim that the problem with many programming languages 
>is that they enforce too many restrictions on the programmer and 
>his code.  Thus, given a problem, the programmer first has to map 
>that problem onto what is permitted by the language.  The less 
>flexibility the language offers, the more difficult this mapping 
>process can be for more problems.  Perl is supposed to reduce the 
>amount of mapping that a programmer has to do before implementation.

The only languages that have taken this idea seriously is the Lisp
family, which have macro systems that basically allow the programmer 
to extend the language. In short: the Lisps have an extensible syntax,
where Perl just has an extensive syntax. 

Python (and Ruby, for that matter) remind me more of Smalltalk than
anything else -- ie, they have extensive libraries, a community with a
love of elegance but not of purity[*], and a greater focus on practical
applications than on academic research. 

The differences are that there's less of a closed-world attitude in
Python than in the ST world, though -- Python can be used for sysadmin
work where ST (even Squeak) wouldn't be a natural fit. But on the flip
side Smalltalk has always had a stronger focus on UI and dev
environments.  This is changing, though -- the ST culture is beginning
to value the ability to deal with the world outside the system image,
and Guido's latest enthusiasm is "computer programming for everybody"
(which is a very ST-ish sort of idea).


Neel



From ferraris at mdmlab.mi.infm.it  Thu Nov 11 03:38:45 1999
From: ferraris at mdmlab.mi.infm.it (Sandro Ferrari)
Date: Thu, 11 Nov 1999 09:38:45 +0100
Subject: How do I compile NumPy on Linux?
Message-ID: <382A8094.3354DE70@mdmlab.mi.infm.it>

I downloaded the LLNLDistribution.tgz file and followed the installation
instruction:
cd Numerical
./makethis.py
but I get the message:
make: Makefile.pre.in: No such file or directory
make: *** No rule to make target `Makefile.pre.in'.  Stop.
make: *** No targets.  Stop.
In the documentation isn't mentioned anything about it.
Any help?
thanks
Sandro




From emile at fenx.com  Tue Nov  9 23:15:15 1999
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 9 Nov 1999 20:15:15 -0800
Subject: __str__ vs. __repr__
References: <000401bf25b9$88ea57e0$2e2d153f@tim> <199911040157.UAA11908@eric.cnri.reston.va.us> <046101bf26d4$2dd1a720$01ffffc0@worldnet.att.net> <m23dulmt3d.fsf@mycroft.actrix.gen.nz>
Message-ID: <004b01bf2b32$31c7cc80$01ffffc0@worldnet.att.net>

Neel,

Ah, but you didn't disagree!  ;-)  Testing certainly is useful, and
for the tester it *is* work, but that's code that is not in production,
which is when the code *is* working.

I quote now from Paul Foley <mycroft at actrix.gen.nz>, who e-mailed me
a sample interactive session using a function describe():

>
> You mean something like this:
>
> >>> x="Testing"
> >>> describe(x)
> A string containing 7 characters.
> >>> import sys
> >>> describe(sys)
> The `sys' module.
> The module has no documentation string.
>
> It exports the following functions: ['setcheckinterval', 'exc_info',
'exit', 'setprofile', 'settrace', 'getrefcount']
>
> It exports the following variables: ['platform', 'stderr', 'version',
'builtin_module_names', 'modules', 'stdin', 'exec_prefix', 'copyright',
'executable', 'exc_type', 'ps1', 'ps2', 'path', 'maxint', 'prefix',
'argv', 'stdout']
> >>> describe(sys.exit)
> The built-in function `exit'.
> >>> describe(sys.stdin)
> A file open for reading on `<stdin>' (a TTY device)
> The underlying file descriptor is 0.
> The current file position is 0.

This is what I had in mind.  To further enhance the utility of
interactive
mode by providing more information to the programmer.  Certainly __str__
and __repr__ are The Wrong Way (tm) to do this, and describe() much
better.
Any and everything more that describe() could do should be added to it.
Add
-v, -vv, -vvv options as appropriate.  Show the source entry line.  Hey,
show
the whole routine. ;-0  (Let's see... get line number from bytecode,
read
source for line, note indent, display until match indent... is that it?)

is-what-we-do-work-ly yr's

--

Emile van Sebille
emile at fenx.com
-------------------


Neel Krishnaswami <neelk at brick.cswv.com> wrote in message
news:slrn82hhtv.ppi.neelk at brick.cswv.com...
> Emile van Sebille <emile at fenx.com> wrote:
> >
> >It seems to me that issues arise when using the interactive
interpreter.
> >I suspect that little useful work is done in this mode, and is used
> >predominately for quick testing while creating python programs.
>
> I'd like to politely disagree . Combined with Tim Peters and Barry
> Warsaw's emacs py-mode, Python's interactive mode is the single most
> useful feature of the language for me. With it, I can test each and
> every method, loop, and statement as I code. This does wonders for my
> understanding of my code. And not incidentally, reliability goes way
> up as well.
>
> I learned Perl before Python, but my Python skill surpassed my Perl
> skill after a much smaller time investment. I lay this pretty much
> wholely at the feet of the interactive interpreter -- I could find out
> the answers to puzzles without a tedious edit-save-run cycle.
>
>
> Neel
>
> --
> http://www.python.org/mailman/listinfo/python-list
>
>





From parkw at better.net  Mon Nov 29 17:07:05 1999
From: parkw at better.net (William Park)
Date: Mon, 29 Nov 1999 17:07:05 -0500
Subject: Htmlizing text
In-Reply-To: <943909426snz@vision25.demon.co.uk>
References: <943909426snz@vision25.demon.co.uk>
Message-ID: <19991129170705.A314@better.net>

On Mon, Nov 29, 1999 at 09:03:46PM +0000, Phil Hunt wrote:
> 
> Is there a function in the standard Python library to HTML-ize text,
> i.e. to replace 'a > b & c < d' with 'a &gt; b &amp; c &lt; d'?
> 
> (I realize this can be done with regular expressions, but is there
> an easy way?)
> 
> -- 
> Phil Hunt - - - phil at comuno.com

The following function replaces [<&>] unconditionally.
    def escapeall(s):
	s = string.replace(s, '&', '&amp;')		# do this first
	s = string.replace(s, '<', '&lt;')
	s = string.replace(s, '>', '&gt;')
	return s
I used this in my 'HTMLtag' module which generates HTML tags.

William.



From doughellmann at home.com  Tue Nov 30 07:12:53 1999
From: doughellmann at home.com (Doug Hellmann)
Date: Tue, 30 Nov 1999 12:12:53 GMT
Subject: bug in tkinter canvas event handling?
References: <38429BB0.1BFFB54B@sage.att.com>
Message-ID: <3843C0A0.94990E64@home.com>


"Garrett G. Hodgson" wrote:
> 
> ok, i'm back with another tk problem.
> what i want to do is attach a callback to a shape in a canvas
> that will tell me which domain level object was picked.
> 
> seems simple, but i don't get what i expect.  in particular,
> if i press a mouse button while pointing to one object, then drag
> to another and release, the release event callback tells me i
> was pointing to the first object.  i've distilled it down to a
> small test case, attached below.
> 
> this seems like a bug to me, but experience has shown that it's
> probably a misconception on my part about how the event handling
> is supposed to work.

This behavior surprised me at first, too.  It's as though the first
canvas object is grabbing the pointer so that all events go to it.  I
got around the problem by looking for the x,y coords of the release
event and then finding the object(s) under that point on the canvas.

Doug



From claudius at catlover.com  Tue Nov  9 23:36:34 1999
From: claudius at catlover.com (claudius at catlover.com)
Date: Wed, 10 Nov 1999 04:36:34 GMT
Subject: Java vs. Python - Serialization Bakeoff
References: <99110918085909.01567@quadra.teleo.net>
Message-ID: <mD6W3.25292$Ec1.280622@news1.frmt1.sfba.home.com>

Patrick Phalen <python-list at teleo.net> says:
>http://starship.python.net/crew/gandalf/DNET/serial/jp-serial.html
>contains an report on tests comparing Java's and Python's data
>marshalling, using various criteria. Numerous interesting charts and
>graphs are presented.
>
>Since this was published in 1996 (Python 1.4), I wonder if anyone
>expert in this subject has any comment on whether the intervening years
>would have changed the conclusions substantively -- namely that Python
>is, on average, faster, simpler, and more flexible.

Hmm?  I get 'not found' (and I'm very interested in reading your findings.)



From aa8vb at yahoo.com  Mon Nov  1 10:58:46 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Mon, 1 Nov 1999 10:58:46 -0500
Subject: eff-bot Python book ... revisited
In-Reply-To: <00fd01bf23cb$e4b6ed30$f29b12c2@secret.pythonware.com>
References: <7va9hf$ngi$1@marcie.fissiontech.com> <19991029081909.B106968704@vislab.epa.gov> <00fd01bf23cb$e4b6ed30$f29b12c2@secret.pythonware.com>
Message-ID: <19991101105846.A133876404@vislab.epa.gov>

Fredrik Lundh:
 |for now, you get a PDF file, which can only be viewed/printed from a
 |Windows box.
...
 |<flame note="not directed to randall, and shouldn't be read by anyone
 |made of flammable material">
 |
 |and before even more people start whining about me "betraying Python in
 |itself, and especially its Unix roots," note that 1) this book wouldn't
 |exist if it hadn't been for the eMatter initiative, 2) I only wrote the
 |book, not the distribution software, and 3) Python was originally written
 |on a Mac.

I'm sorry you might have interpreted my post that way.  I think it is good
that Python is getting so much coverage in the media, both printed an
on-line (with more coming soon).  This is a good first-start by eMatter!

As a consumer though, I feel incumbant to express my desire for a product
which is very interesting but does not fit my needs.  A consumer-aware
company will gear their product to demand, so they need feedback.  I rarely
use MSWindows and do no development in it.  So an on-line, searchable
reference tied to Windows doesn't exactly work.

    I'm sure I could steal the docs out from under their viewing shell, but
I'm not sure all UNIX/Linux users could.  For personal use this sounds like
a reasonable work-around, and I'm tempted.  I assume this doesn't violate
their usage license?

-- 
Randall Hopper
aa8vb at yahoo.com



From news at dorb.com  Fri Nov 12 11:21:29 1999
From: news at dorb.com (Darrell)
Date: Fri, 12 Nov 1999 11:21:29 -0500
Subject: .dll and .exe file version informaton
References: <816010E2456BD111A48700805FBBE2EEFDF793@ex-quebec-u1.baan.com> <80ha4a$pmg$1@nnrp1.deja.com>
Message-ID: <C6XW3.10409$Xd2.618639@typhoon.nyroc.rr.com>

Given the win32 calls it wouldn't be too hard to wrap them up in an
extension module.

--
--Darrell
<edwardm_painewebber at my-deja.com> wrote in message
news:80ha4a$pmg$1 at nnrp1.deja.com...
> It looks like they are not available yet....
> In article <816010E2456BD111A48700805FBBE2EEFDF793 at ex-quebec-
> u1.baan.com>,
>   Gaetan Corneau <Gaetan_Corneau at baan.com> wrote:
> > > go-ahead-just-*try*-to-write-a-program-to-do-that<wink>-ly y'rs  -
> tim
> >
> > If you really want to do it in a program, take a look at the following
> > functions in the help:
> > GetFileVersionInfo
> > VerQueryValue
> > QueryField
> >






From chapman at bioreason.com  Tue Nov 16 19:56:13 1999
From: chapman at bioreason.com (Mitch Chapman)
Date: Tue, 16 Nov 1999 17:56:13 -0700
Subject: Beginning programming with Python
References: <sng0ychgwt.fsf@motorola.com> <80p9ds$ppo$1@newshost.accu.uu.nl> <asteroidal-80psfo/INN-2.2.1/bat@broadway.news.is-europe.net> <38317787.19AFDAF5@compaq.com> <80s22o$o0l$1@nntp9.atl.mindspring.net>
Message-ID: <3831FD2D.2FF2087A@bioreason.com>

Aahz Maruch wrote:
> 
> In article <38317787.19AFDAF5 at compaq.com>,
> Greg Ewing  <greg.ewing at compaq.com> wrote:
> ...
> >Well, not quite - you can forget the "self" in method
> >declarations fairly easily...
> 
> I don't seem to have that problem; my fingers have learned to
> automatically type "(self" whenever "def" is indented.  ;-)

If you use [X]Emacs you can use python-mode functions
to generate function and method templates -- including 
automatic insertion of the leading "(self" on method defs:

http://starship.python.net/crew/mchapman/pycc.el

This is old code, and word is that use of template.el would have 
greatly simplified it.  But it works for me.

--
Mitch Chapman
chapman at bioreason.com



From bernhard at alpha1.csd.uwm.edu  Mon Nov 29 13:42:27 1999
From: bernhard at alpha1.csd.uwm.edu (Bernhard Reiter)
Date: 29 Nov 1999 18:42:27 GMT
Subject: Pythonwin Reset
References: <01bf38df$8127e0c0$3a42c3d1@Cyberus.cyberus.ca> <_X004.7432$VA6.38559@news-server.bigpond.net.au> <01bf39bd$a28831a0$4155c3d1@Cyberus.cyberus.ca> <81rmh5$9e$1@hub.org> <01bf3a8c$3f855b80$2241c3d1@Cyberus.cyberus.ca>
Message-ID: <slrn845i8j.178.bernhard@alpha1.csd.uwm.edu>

On Mon, 29 Nov 1999 17:08:40 GMT, Gordon Williams <g_will at cyberus.ca> wrote:

>It now works in that it eliminates all the names, but it doesnt quite fix
>my problem.  When I make a GUI with wxPython (using Pythonwin IDE) and I
>get an exception, the next time I try and run the program after fixing the
>mistake, I get another exception saying One wxApps process only!  Something
>was left behind.  The only way I have found to clear the problem was to
>close Pythonwin and then restart it and reload the file.

There is no fix AFAIK, but you can debug your little program in a
different way.

Add another method to the windows exploder, something like
debug and add the option -i before the "%1" where python is called.
Then you can rightclick on the python file and select debugging.
You will see the error messages when the programs runs in 
the DOS window.

(Close it with Control-Z)

Should be a FAQ somewhere.
	Bernhard
-- 
Research Assistant, Geog Dept UM-Milwaukee, USA.  (www.uwm.edu/~bernhard)
Free Software Projects and Consulting 		         (intevation.net)  
Association for a Free Informational Infrastructure            (ffii.org)



From tcdoe at pitt.edu  Thu Nov  4 15:04:27 1999
From: tcdoe at pitt.edu (tdoe)
Date: Thu, 4 Nov 1999 15:04:27 -0500
Subject: file i/o !!! help
Message-ID: <7vsolo$aer$1@usenet01.srv.cis.pitt.edu>

Hi,
im very new to using python, and im having a TERRIBLE time trying to find
some example scripts for FILE I/O !!!!
all i want to do is read a "comma" or "space" delimited list of numbers from
a file such as:

1 2 3 4
5 6 7 8
9 10 11 12

and input them into a floating point list variable.

well, this is probably a "cinch", but
for me this has been tough.

any help very appriciated.
thanks,
todd
tcdoe at pitt.edu






From a.eyre at optichrome.com  Mon Nov 22 11:34:08 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Mon, 22 Nov 1999 16:34:08 -0000
Subject: Getting PythonWin debugger working with an embedded app.
In-Reply-To: <11A17AA2B9EAD111BCEA00A0C9B41793034AAF00@molach.origin.ea.com>
Message-ID: <000a01bf3507$6600b9b0$3acbd9c2@peridot.optichrome.com>

> [snip]
> We are using the debug version of the Python DLL for Windows.

There's the clue.

> [snip]
> ImportError: No module named win32ui

If you read through the Python source, you will find that the
debug Python DLL requires all extension modules to end with '_d'
(e.g. win32ui.pyd must be named win32ui_d.pyd). There is a good
reason for this. It would appear that mixing DLLs built with
MSVCRT.DLL and MSVCRTD.DLL causes problems. Unfortunately this
means you can't just rename the module. You must build a debug
version from the source. I know it's a pain in the arse,
especially for the modules that need SWIG in order to build,
and/or you only have MSVC5 (this requires a few changes to
defined symbols.)

I would like to suggest to Mark Hammond, if possible, to include
debug versions of the win32all modules in the distribution, maybe
as an optional component in the installer.

I would also like to suggest the same to Guido van Rossum in the
Python installer.

Just including the release version is fine for developers in pure
Python, but anyone embedding or extending Python will hit the
same problem sooner or later.

> Any advice or suggestions would be truly welcome.  Getting the
> debugger working would be a huge step forward for us.

Let me know if you have any success, as I managed to get this far,
but hit a problem later on.

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From n89553 at squid.upb.de  Fri Nov  5 10:43:04 1999
From: n89553 at squid.upb.de (Mirko Liss)
Date: Fri, 05 Nov 1999 16:43:04 +0100
Subject: Bug: whichdb.py of python 1.5.2 at suse linux 6.2
Message-ID: <3822FB08.EC0CD544@squid.upb.de>

Hi,

the distribution SuSE 6.2 (linux at ix86) comes with
a flavour of bsd-db that uses the magic key 0 (zero) in
its databases. This feature makes it impossible to open 
existing shelves.

This can be fixed by editing /usr/lib/python/whichdb.py :

    # Check for BSD hash
# patch for suse 6.2 , mirko liss
#    if magic in (0x00061561, 0x61150600):
    if magic in (0x00061561, 0x61150600, 0):
        return "dbhash"
                 

regards,
 Mirko
 

-- 
M Li?, <n89553 at squid.upb.de>



From skip at mojam.com  Mon Nov 29 16:37:01 1999
From: skip at mojam.com (Skip Montanaro)
Date: Mon, 29 Nov 1999 15:37:01 -0600 (CST)
Subject: 50% Python content :)   (Distributed computing)
In-Reply-To: <816010E2456BD111A48700805FBBE2EEFDF862@ex-quebec-u1.baan.com>
References: <816010E2456BD111A48700805FBBE2EEFDF862@ex-quebec-u1.baan.com>
Message-ID: <14402.61949.484621.155114@dolphin.mojam.com>

    Gaetan> I would like to explore distributed programming, and I think
    Gaetan> Python would be the right tool for that. I don't want to produce
    Gaetan> industrial-strenght code right now, but learn the basics. 

There are a number of different models of distributed computing available.
You might find playing around with the eff-bot's xmlrpclib module (check
www.pythonware.com) a relatively painless way to experiment with the remote
procedure call model.  I saw something posted recently about an
implementation of a tuple space interface for Python as well.  Check out

    http://www.snurgle.org/~pybrenda/

You'll find that a bit more complicated to get started with, but may in the
long run provide a better computational model (depends on what you're really
after).

Finally, there is a Python interface to PVM.  Check out

    http://www.flyn.org/

to-play-is-to-be-happy-ly y'rs,

Skip Montanaro | http://www.mojam.com/
skip at mojam.com | http://www.musi-cal.com/
847-971-7098   | Python: Programming the way Guido indented...



From fdrake at acm.org  Thu Nov 11 09:45:08 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Thu, 11 Nov 1999 09:45:08 -0500 (EST)
Subject: .html -> .ps ?
In-Reply-To: <19991111080506.A1007@better.net>
References: <19991111080506.A1007@better.net>
Message-ID: <14378.54900.714487.334690@weyr.cnri.reston.va.us>

William Park writes:
 > 1.  Does anyone have experience with 'html2ps', 'html2tex', and
 >     'html2latex' programs?  Or, is there any other program that will do
 >     the conversion in Python or otherwise?  

  There's an html2ps.py in the Grail distribution; if your HTML
doesn't involve tables that might just do it fo you.  You can even
drive it from Python by using Grail's "printing" package.
  Unfortunately, most HTML these days seems to involve at least tables 
nested 12-deep.  ;-(


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From hei at adtranzsig.de  Thu Nov 25 04:52:11 1999
From: hei at adtranzsig.de (Dirk-Ulrich Heise)
Date: Thu, 25 Nov 1999 10:52:11 +0100
Subject: Python complaints
References: <XFMail.991124113221.mikael@isy.liu.se> <81gj3i$fe3$1@mach.vub.ac.be> <005201bf367e$4901fef0$f29b12c2@secret.pythonware.com> <81gpd5$332$1@mach.vub.ac.be>
Message-ID: <943526916.548103@news.adtranzsig.de>

Thomas Hamelryck schrieb in Nachricht <81gpd5$332$1 at mach.vub.ac.be>...
>: > indentation or not. In many cases, using a variable indentation
(depending
>: > on what construct you are using or the length of a line) leads to code
>: > that is visuably much more appealing.

Once i saw a C program indented like a poem (centered lines).
Visually, it was very appealing. I was glad i didn't have to understand it,
though.

--
Dipl.Inform. Dirk-Ulrich Heise
hei at adtranzsig.de
dheise at debitel.net





From tismer at appliedbiometrics.com  Wed Nov 17 09:17:07 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Wed, 17 Nov 1999 15:17:07 +0100
Subject: ANNOUNCE: a few little things
References: <slrn833s87.elm.nascheme@adler.dynodns.net>
Message-ID: <3832B8E3.3298AB47@appliedbiometrics.com>


Neil Schemenauer wrote:
> 
> I spent a bit of time today and updated my Python web site.  The
> highlights are:
> 
> py152_sa.zip
> ------------
> 
> A Stand-alone version of Python for Windows. It includes Python
> 1.5.2, Tcl, zlib and bsddb.  It fits on a floppy and can be
> extracted and run from any location.  Just add your application
> code.

Just a question here:
I did a stand-alone version too a while ago, but didn't publish
it for the following reason:
If a user happens to import *any* .pyd module which is built
for Windows, this pyd will try to load python15.dll since they
all link against it. Did you provide something which at least
prevends this problem?
One possible solution would be a stub python15.dll just to be safe,
which itself is linked against you .exe file. This works as far as
I remember, but then you are at two files again, and the question
remains why this is better than to have what we now have?

cheers - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From python-list at teleo.net  Mon Nov 29 12:38:26 1999
From: python-list at teleo.net (Patrick Phalen)
Date: Mon, 29 Nov 1999 09:38:26 -0800
Subject: Running apps with Zope 2?
In-Reply-To: <3842B384.5276F337@gmd.de>
References: <80v5g8$5a6$1@mulga.cs.mu.OZ.AU> <383E7D7A.1B48C7F1@gmd.de> <3842B384.5276F337@gmd.de>
Message-ID: <99112909422700.03205@quadra.teleo.net>

[Peter K., on Mon, 29 Nov 1999]

:: Anyone with better ideas?

1. Zope won't publish objects without a doc string (as your error msg.
indicates).

2. Subscribe to the Zope mailing list (http://zope.org), where you'll
find a large community of Zope-aware citizens willing to help.





From jblazi at netsurf.de  Tue Nov 16 13:02:08 1999
From: jblazi at netsurf.de (Janos Blazi)
Date: Tue, 16 Nov 1999 19:02:08 +0100
Subject: Beginning programming with Python
References: <sng0ychgwt.fsf@motorola.com> <382B3A53.5816B2@callware.com> <382B3D79.74CD3E0F@email.sps.mot.com> <80p9ds$ppo$1@newshost.accu.uu.nl> <asteroidal-80psfo/INN-2.2.1/bat@broadway.news.is-europe.net> <38312D55.A36EB56D@iqsoft.hu> <antigorite-80rol2/INN-2.2.1/attentive@broadway.news.is-europe.net> <383195C1.826B3342@iqsoft.hu>
Message-ID: <allege-80s674/INN-2.2.1/apply@broadway.news.is-europe.net>

> As far as I understood, you are absolutely sure that Python is a perfect
> language, I have to trust in it without any hesitation?

I said that I have not had any experience using Python in my teaching yet
and that I should like to try to use it first before I judge it. Do these
words and "absolutely sure" mean the same thing to you? Or are you
practising your English?

Janos B.





From andres at corrada.com  Sun Nov 28 16:26:58 1999
From: andres at corrada.com (andres)
Date: Sun, 28 Nov 1999 16:26:58 -0500
Subject: Parsing text
References: <g1e04.1$O8.108@news.inreach.com>
Message-ID: <38419E22.7F6530EF@corrada.com>

string.find(...) should do the trick



From gerrit.holl at pobox.com  Sun Nov 14 13:41:18 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Sun, 14 Nov 1999 19:41:18 +0100
Subject: telnetten vanuit een script
In-Reply-To: <19991112201259.A6212@optiplex.palga.uucp>
References: <Pine.LNX.4.10.9911121516030.28857-100000@semios.c75.net> <19991112201259.A6212@optiplex.palga.uucp>
Message-ID: <19991114194118.A30062@optiplex.palga.uucp>

Gerrit Holl wrote:
> Bert Frederiks schreef:
> > Ik wil graag vanuit een script telnetten, dan een paar dingen doen en weer
> > uitloggen en ondertussen de uitvoer opvangen in een bestandje. Dat werkt
> > prima met pop3 op poort 110 maar op poort 23 lukt het me niet om in te
> > loggen. Het probleem zit hem aan invoerkant want dit werkt wel:


Sorry, wrong mailinglist!

regards,
Gerrit.



From sblakey at home.com  Tue Nov 30 18:46:10 1999
From: sblakey at home.com (Sean Blakey)
Date: Tue, 30 Nov 1999 23:46:10 GMT
Subject: What's the canonical vi setting for Python indentation
In-Reply-To: <38445B6A.A1283D2E@cs.mu.oz.au>
References: <38445B6A.A1283D2E@cs.mu.oz.au>
Message-ID: <Pine.LNX.4.20.9911301540450.30390-100000@seanb.sttln1.wa.home.com>

I don't know about canonical, but I have the following in my .vimrc
set expandtab	"Turn's tabs into spaces
set hardtabs=4	"Make tabs 4 spaces wide
set tabstop=4
set shiftwidth=4	"For use with << and >>

On Wed, 1 Dec 1999, Chris WRIGHT wrote:

> I've tried to find out the "OK" indentation settings for python with vi,
> and a troll through deja-news found some hints (like don't do indent =
> 4). Yet I seem to also remember some advice to the contrary?? I want
> indentation to be 4 characters, and readable by/ acceptable to emacs
> python-mode, and to pass tabnanny...
> 
> I think that this should go in the faq...(Couldn't find it from
> www.python.org search)
> 
> 
> cheers
> and thanks
> 
> chris wright
> 

-- 
http://students.washington.edu/sblakey/
(206)297-7123
Public PGP key available at http://students.washington.edu/sblakey/public.asc






From Dan at Grassi.com  Sun Nov 28 14:30:04 1999
From: Dan at Grassi.com (Dan Grassi)
Date: Sun, 28 Nov 1999 14:30:04 -0500
Subject: IDLE & Linux
Message-ID: <B466ECEC.F8E%Dan@Grassi.com>

I am trying to get IDLE to run.  The docs say that it is automatically
installed but that is just not true.  There is a /usr/bin/idle but when it
runs it quits complaining that it can't find
/usr/lib/python1.5/site-packages/idle/idle.py

Of course I can not find where to get the missing files.  The rpm package I
have does not include this package.

I am running RH 6.1 and do have TCL/TK installed.

Any idea of what to do next?

Dan




From geoff at gnaa.net  Sat Nov 13 19:06:50 1999
From: geoff at gnaa.net (geoff nordli)
Date: Sat, 13 Nov 1999 16:06:50 -0800
Subject: NT admin scripts with python
In-Reply-To: <jEmX3.3257$Xo.17395@news-server.bigpond.net.au>
Message-ID: <000201bf2e34$26d93750$020a10ac@p350.gnaa.net>

Excellent.  I will give it a go.

Does anyone know if there is a reservoir of python scripts on the net.  Kind
of
like Perl's CPAN

thanks,

geoff

-----Original Message-----
From:	python-list-admin at python.org [mailto:python-list-admin at python.org] On
Behalf Of Mark Hammond
Sent:	Saturday, November 13, 1999 3:39 PM
To:	python-list at python.org
Subject:	Re: NT admin scripts with python

Yep - the win32net module is all you need!  Included in the win32all
extensions...

Mark.

geoff nordli wrote in message
<000101bf2e22$7060a190$020a10ac at p350.gnaa.net>...
>Does python have a module that would allow a person to write
>scripts on NT Server to lets say:  create users, map drives,
>modify registry, create shares, and so on.
>
>thanks,
>
>Geoff Nordli
>
>
>



--
http://www.python.org/mailman/listinfo/python-list




From pmoran at nas.nasa.gov  Wed Nov  3 18:08:21 1999
From: pmoran at nas.nasa.gov (Patrick J. Moran)
Date: 3 Nov 1999 23:08:21 GMT
Subject: NT DLL modules and C++
References: <002701bf2609$38622320$3acbd9c2@peridot.optichrome.com>
Message-ID: <7vqf95$5rg$1@sun500.nas.nasa.gov>

In article <002701bf2609$38622320$3acbd9c2 at peridot.optichrome.com>, "Adrian Eyre" <a.eyre at optichrome.com> writes:
|> > How do I get Python to call the initialization under NT?
|> 
|> This should do the trick (for module spam.dll):
|> ...

I should clarify a bit.  I have an initFOO() call for my FOO module, and it gets called
just fine.  I can manually call "placement new" constructors on the static data members
of my classes to initialize them, but it's awkward and a hack.  It's also not a complete
solution because there may be other static objects that belong to libraries whose
implementation I'm not familiar with, e.g., iostreams.  There's got to be a better way.

My understanding is that there is an entry point, _DllMainCRTStartup, that the linker
generates that calls the necessary C++ constructors, and then calls DllMain.  So, the
good news is that I could call _DllMinCRTStartup and the appropriate construction code
would be called, assuming I figure out the arguments required by _DllMainCRTStartup.
The bad news is that _DllMainCRTStartup calls DllMain, the entry point provided by
the user (or the Python library, I presume).  DllMain in turn calls my initFOO (again,
I presume) so I would be in an infinite loop.  Of course I could think of hacks to prevent
the infinite looping.  Still, there's got to be a better way.

The solution that I'm looking for is to get Python to call the constructor code when
it does the dynamic linking.  I think this means calling _DllMainCRTStartup rather than
DllMain.  I've tried looking through the doc in the PC directory, but it's still
unclear to me.  That's what I was looking for help with.

Thanks.

Pat






From Friedrich.Dominicus at inka.de  Mon Nov 29 02:35:29 1999
From: Friedrich.Dominicus at inka.de (Friedrich Dominicus)
Date: Mon, 29 Nov 1999 08:35:29 +0100
Subject: Recursion?
References: <38421570.9C3D9AEC@videotron.ca>
Message-ID: <38422CC1.7DB0B37B@inka.de>

Pierre-Yves Cloutier wrote:
> 
> Can I use recursion in python?

Yes.

Regards
Friedrich



From kent at tiamat.goathill.org  Tue Nov 30 17:03:06 1999
From: kent at tiamat.goathill.org (Kent Polk)
Date: 30 Nov 1999 22:03:06 GMT
Subject: Missing_cmp (was Re: '==' vs. 'is' behavior)
References: <382C2C6A.64A8F2C2@theriver.com> <14380.12036.197083.409565@weyr.cnri.reston.va.us> <8E7F86267duncanrcpcouk@news.rmplc.co.uk> <oqemd9dk2g.fsf@titan.progiciels-bpi.ca> <015301bf3b0a$685a3fc0$f29b12c2@secret.pythonware.com> <oqg0xnapew.fsf@titan.progiciels-bpi.ca> <3dbt8bhhje.fsf@amarok.cnri.reston.va.us>
Message-ID: <943998076.939362@fezzik.endicor.com>

On 30 Nov 1999 15:05:09 -0500, Andrew M. Kuchling wrote:

) At the time I thought the Python comparison was a bit
>odd, and didn't really see the point.  Python is more like a
>collection of tiny interacting modules; no one module is very
>complicated, but put them together and they intermesh and interweave
>in a beautiful way over time. [1] Over time the modules fall together
>in slightly different patterns, and, though the difference between
>iterations is slight, the resulting shift can completely change the
>effect of the whole piece.

Speaking of weaving Python module comparisons together...

The module Missing, which provides a placeholder mechanism for
missing data (one of Jim's ExtensionModules for Zope) has some
interesting cmp behavior:

>>> import Missing
>>> b=[5,Missing.Value,0,-3,'c',Missing.Value,0,4,'a',Missing.Value,
  'b',Missing.Value,6]
>>> b.sort()
>>> b
[-3, 5, Missing.Value, 0, Missing.Value, 0, 4, Missing.Value,
   Missing.Value, 6, 'a', 'b', 'c']

As you can see, inserting a Missing.Value before a zero in a list
causes problems with numeric compares.

Note that Missing.Values currently evaluate to 0, and the current
compare method for Missing objects is:

static int
Missing_cmp(Missing *m1, Missing *m2)
{
  return m1->ob_type != m2->ob_type ;
}

which always returns 1 on failure. This leads to evaluating 5 as
less than Missing.Value under certain circumstances.  A better
solution would be to always return -1 on failure so that Missing.Value
evaluates to less than any object greater than zero, as in:

[-3, Missing.Value, Missing.Value, Missing.Value, 0, 0, 4, 5, 5,
'a', 'c']

My question is whether this is the desired behavior or not...
Should Missing.Value really be ordered as equal to zero or should
it be less than other object types?  I suspect that if it evaluates
to a non-zero value, problems might ensue with how people handle
no data, so I suppose this should be sufficient behavior. However,
is there any way for Missing.Value to evaluate to zero, yet have
the compare function order Missing.Values as less than other object
types?  (ha)

Thanks



From philh at vision25.demon.co.uk  Wed Nov 24 16:10:52 1999
From: philh at vision25.demon.co.uk (Phil Hunt)
Date: Wed, 24 Nov 99 21:10:52 GMT
Subject: RDF/RSS
References: <943225647snz@vision25.demon.co.uk>
Message-ID: <943477852snz@vision25.demon.co.uk>

In article <943225647snz at vision25.demon.co.uk>
           philh at vision25.demon.co.uk "Phil Hunt" writes:
> As part of my Comuno website, I am creating an RDF file so that other
> sites can link to my changing content. I have written some software
> to handle RDF (written in python of course), and will be releasing it
> under an open source license in the not-too-distant future, i.e.
> when I have got it to produce HTML.

I was going to call this Python Site Summary, or "pyss" for short,
but then decided perhaps that wasn't appropriate. Looks like I'll 
have to think up a better name. (Or perhaps someone at the UK
Python London meeting this Saturday[1] will come up with one.)


[1] 7:30 pm, at Bier Rex in Putney, with is located on Putney High
Street, 100 m south of Putney bridge, opposite the cinema. All are 
welcome. We'll have a plastic snake with us, so you can tell who
we are. Email me for further details -- see address in my .sig.

-- 
Phil Hunt - - - phil at comuno.com
Want to see the FAQs for your favourite newsgroups? Read them on
the web with Comuno's FAQ server at http://www.comuno.com/faq/ 




From thomas at bibsyst.no  Mon Nov  1 04:33:56 1999
From: thomas at bibsyst.no (Thomas Weholt)
Date: Mon, 01 Nov 1999 10:33:56 +0100
Subject: Need Berkley DB!!
Message-ID: <381D5E84.98931682@bibsyst.no>

Hi,

After I spent most of my weekend trying to get Berkley DB to work with
Python on my Linux ( Suse 6.1) machine, I`ve now given up completly.
I`ve tried to install DB.1.85. and DB.2.7.7, neither worked, either
crashing under installation ( which of course, is not Python
related ) or when I tried to test the installation with the examples in
the documentation ( the wrapper for the 2.7.5-module listed at
www.python.org, under modules, databases ).

Puh!! I really need some database, cuz` dumbdbm seem to be ineffecient
( both on size and speed ) when I did some serious testing. My question
is ; is there another database-module/dist. I can use with Python which
is platform-independant ( and perhaps easier to get to work ) or is
there a Python version available with built-in Berkley DB-support ( libs
and all, no more installation needed )??

I know the win32all.exe stuff and Python for Windows seem to have these
things built in but I really don`t want to go back to programming in
Windows again, even if it is using Python.

Any ideas?! I really need help on this one.

Best regards,
Thomas Weholt




From gmcm at hypernet.com  Wed Nov  3 22:03:21 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 3 Nov 1999 22:03:21 -0500
Subject: How does rexec work?
In-Reply-To: <3820C73B.9278E532@mincom.com>
Message-ID: <1270437901-37926349@hypernet.com>

John Farrell writes:

> I'm confused about rexec and how it works. 

Aren't we all.

> Consider the following
> code: --- import rexec, traceback
> 
> renv1 = rexec.RExec()
> renv1.r_exec('import string')
> print renv1.r_eval("string.upper('This is a test')")
> 
> try:
>     renv2 = rexec.RExec()
>     renv2.r_import('string')
>     print renv2.r_eval("string.upper('This is a test')")
> except:
>     traceback.print_exc()
> ----
> The things I don't understand are:
> 
>  (a) In the renv1 example, where does the name string get
>  preserved
> to be available in the eval? The relevant code from rexec is
> included below, but I don't see how the self.modules dict gets
> into the __main__ namespace.

It doesn't. "renv1.r_exec('import string')" binds the module 
string to the name "string" in renv1's __main__. __main__ is a 
mutable object, stored in renv1.modules. So it's available to 
your "renv1.r_eval(...)".

Looks like renv1.modules plays the same role here that 
sys.modules plays normally. That is, if a module has been 
imported somewhere, it's in sys.modules. You still need to 
import it in other modules, but those imports are nearly free.

>  (b) Why doesn't r_import work like r_exec('import ...')? What's
>  it
> supposed to do then?

r_import is like __import__, not like the keyword import. It 
returns the module.

    mod = renv2.r_import('string')
    renv2.modules['__main__'].string = mod

> I guess part of my problem is that I am not 100% familiar with
> the Python namespace fiddling stuff yet. Thanks for any help!

That'll teach you to dive in the deep end!

- Gordon



From mikael at isy.liu.se  Fri Nov  5 04:10:32 1999
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Fri, 05 Nov 1999 11:10:32 +0200 (MET)
Subject: ~ 1 = -2 ???
In-Reply-To: <m2vh7hk1ys.fsf@netmeg.net>
Message-ID: <XFMail.991105111032.mikael@isy.liu.se>

On 05-Nov-99 Les Schaffer wrote:
 >  i am doing some bit operations, and dont understand the following:
 >  
 > >>> ~ 1
 >  -2
 >  
 >  huh???? i would have thought that bitwise invert of 1 would be
 >  
 >  0001 --> 1110 = 13 (base 10)
 >  
 >  what marble am i missing?

Hmmm, I always thought that 1110 binary would be 14 in base 10. Then it
makes sence, since 14 and -2 are equivalent mod 16 (4 bits).

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    05-Nov-99
Time:    11:08:02

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



From python-list at teleo.net  Sat Nov  6 12:04:41 1999
From: python-list at teleo.net (Patrick Phalen)
Date: Sat, 6 Nov 1999 09:04:41 -0800
Subject: making the Python case to my dept.
In-Reply-To: <3823FC43.CF7FBD71@python.net>
References: <9911051740130D.02530@william.jinsa.org> <3823FC43.CF7FBD71@python.net>
Message-ID: <99110609085702.02970@quadra.teleo.net>

[Thomas A. Bryan, on Sat, 06 Nov 1999]:
:: William Ahern wrote:
:: 
:: > Could anybody suggest some substantive info to make the case (or not make it).
:: > Also, the department will need an IDE (Win95/Mac), so if anybody knows where to
:: > find one....

[snip well-reasoned, even-handed answer]

Since this is a FAQ, I'd vote for preserving Thomas' answer somewhere
on the python.org Web site, so people could point to the URL when the
question comes up again.



From eggert at twinsun.com  Sat Nov 13 00:31:29 1999
From: eggert at twinsun.com (Paul Eggert)
Date: 12 Nov 1999 21:31:29 -0800
Subject: Comparison between Python and "Ruby"
References: <1270663667-24391282@hypernet.com>  <381ED7A6.D7D7FB13@compaq.com>  <877lk1wiij.fsf@ev.netlab.co.jp>  <19991102154543.A4469@optiplex.palga.uucp>  <381F9EB4.63FA033F@python.net> <199911031734.JAA15489@hermes.aspectdv.com>
Message-ID: <80it3h$3d2$1@green.twinsun.com>

Jim Althoff <jima at aspectdv.com> writes:

>Absolutely right.  Part of the original dream of the Smalltalk development
>team was to build a programming language/environment that would be so
>easy that anybody -- even kids -- could use it.

Kids?  Kids??  You think _kids_ were the target!?
No way.  The PARC developers were much more ambitious than that.
Smalltalk's design goal was to be so easy
that even _secretaries_ could use it.



From gawron at obop.com.pl  Mon Nov 15 08:24:22 1999
From: gawron at obop.com.pl (=?iso-8859-1?Q?Przemys=B3aw?= G. =?iso-8859-1?Q?Gawro=F1ski?=)
Date: Mon, 15 Nov 1999 14:24:22 +0100
Subject: Binary search
Message-ID: <38300986.E734E9E2@obop.com.pl>

I read somewhere that Python has a build in binary search.

How can I use it ?

--
Przemyslaw G. Gawronski UIN:8358522
mailto:gawronskip at usa.net
mailto:gawron at obop.com.pl





From hildeb at www.stahl.bau.tu-bs.de  Sun Nov 28 10:46:10 1999
From: hildeb at www.stahl.bau.tu-bs.de (Ralf Hildebrandt)
Date: 28 Nov 1999 15:46:10 GMT
Subject: fcntl/ioctl and O_NONBLOCK
References: <slrn840205.sn4.hildeb@stahlw06.stahl.bau.tu-bs.de> <slrn83t5mv.bhk.hildeb@stahlw06.stahl.bau.tu-bs.de> <81ml5t$eer$0@216.39.151.169> <81p32a$m14$1@216.39.151.169> <slrn8405je.q0.hildeb@stahlw06.stahl.bau.tu-bs.de> <38407496$0$222@nntp1.ba.best.com>
Message-ID: <slrn842ji2.o62.hildeb@stahlw06.stahl.bau.tu-bs.de>

On 28 Nov 1999 00:17:26 GMT, Cove Schneider <cove at shell10.ba.best.com> wrote:

>> This is also fixed now, but now the popen3 Object generates about 1 MB
>> of output (which I read using select() and read()), and then it STILL
>> doesn't return. No crash, no exit. And once more, in top it says
>> "sleep" once more. 
>> Any ideas what's going on there?
>
>Humm.. Sounds like your not checking for an EOF..

Good idea. When I read the input file (which I then tochild.write()),
I get only 1000 bytes - althought the file is much longer. Question:
Is it possible to read OVER an EOF? How? 



From gmcm at hypernet.com  Sun Nov 21 14:17:16 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sun, 21 Nov 1999 14:17:16 -0500
Subject: Order of keyword arguments
In-Reply-To: <38382DA7.A03A2023@corrada.com>
Message-ID: <1268910662-35645762@hypernet.com>

Andres Corrada wrote:
> Gordon McMillan wrote:
> > 
> > Ah, but that would be "alphabetical order", not "entry order",
> > and not much use to anybody if it were.
> > 
> 
> I disagree. I think you are confusing entry order with type of
> ordering. entry order could be alphabetical but it could also be
> by time of entry, for example. My point still stands: an ordered
> dictionary would be a useful data structure to have.

Well I'm feeling amiably disagreeable too, so I'll differ with that. 
A "dictionary" in the wider-than-Python sense has an ordering 
to it that is based on it's keys. The order of arguments to a 
function is _not_ an ordering based on the argument names; 
ergo, the existence of an American Heritage Dictionary gets 
you no closer to your goal (in this sense, at least). What 
you're apparently arguing for is an ordered dictionary. What 
you _want_ is an indexed list.

> > Python data structures that would suit your apparent desires:
> >  [ (optionname, optionvalue), ... ]
> 
> Mike Fletcher suggested this also and it is the way I'm getting
> to eat my cake and have it too :-)

bon-apetit-ly y'rs

- Gordon



From phd at phd.russ.ru  Sat Nov 13 09:06:15 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Sat, 13 Nov 1999 14:06:15 +0000 (GMT)
Subject: Production Web Deployment?
In-Reply-To: <382D68D3.46CEFD61@digicool.com>
Message-ID: <Pine.LNX.4.20.9911131403250.3572-100000@fep132.fep.ru>

On Sat, 13 Nov 1999, Paul Everitt wrote:
> Seems nobody in the Zope community is eager to maintain
> an Apache module. :^)

   I've started to write mod_pcgi2. I don't want to use mod_fastcgi, as it
is incompatible with Russian Apache (special patch for Apache to recode
different charsets on the fly). I need virtual hosting, but didn't find a
way to extend AMK's mod_pcgi, so I started to write new module from
scratch. It seems I'll spend quite a lot of time, though :(

Oleg.
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From mhammond at skippinet.com.au  Sun Nov 28 18:55:26 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sun, 28 Nov 1999 23:55:26 GMT
Subject: Python and COM: a bug?
References: <816010E2456BD111A48700805FBBE2EEFDF831@ex-quebec-u1.baan.com> <Puo=OIr6Wof+GYnxsYrc7YwdPjqH@4ax.com>
Message-ID: <Ohj04.8263$VA6.42517@news-server.bigpond.net.au>

Toby Dickenson wrote in message ...

>Not quite. Clients are only allowed to cache the ID for use with that
>one object. DISPIDs are allowed to be different between instances of
>one class, just not during the lifetime of one object.
>
>From what you have described (if I understand correctly) then you
>should be ok????????

Yes - you are correct, and I believe this is indeed the problem.  win32com
needs to be fixed here.

Thanks,

Mark.





From boud at rempt.xs4all.nl  Fri Nov 19 04:45:13 1999
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: Fri, 19 Nov 1999 10:45:13 +0100 (CET)
Subject: Beginning programming with Python
In-Reply-To: <00ef01bf3271$10a8bb90$f29b12c2@secret.pythonware.com>
Message-ID: <Pine.LNX.4.10.9911191043080.2313-100000@essle.valdyas.org>

On Fri, 19 Nov 1999, Fredrik Lundh wrote:

> Boudewijn Rempt <boud at rempt.xs4all.nl> wrote:
> > I tend to forget the colons after function and class definitions and
> > everywhere else and the double == in if's - mainly because, for the life
> > of Brian, I can't see why they are there. It seems to me the language
> > would be just as easily parseable and readable without this bit of
> > interpunction.
> 
> guess you weren't part of the CWI design team -- they
> did test this on humans, you know, and discovered that
> the language wasn't as easily parseable and readable
> without the colons...
> 

You needn't guess about that, I guess ;-). I didn't know there
was a design team for Python, I though it was Guido's invention.
But I still wonder about the double == - if I can't assign in
an if, why insist on ==?


Boudewijn Rempt  | http://denden.conlang.org




From hnowak at cuci.nl  Fri Nov 19 01:15:18 1999
From: hnowak at cuci.nl (Hans Nowak)
Date: Fri, 19 Nov 1999 07:15:18 +0100
Subject: Python in DOS
Message-ID: <199911190617.HAA08663@dionysus.fw.cuci.nl>

On 18 Nov 99, William Park wrote:

> On Thu, Nov 18, 1999 at 06:24:25PM -0500, Curtis, Craig M. wrote:
> > Does anyone know the "approved" way to get rid of the following
> > message:
> > 
> > Could not find platform dependent libraries <exec_prefix>
> > Consider setting $PYTHONHOME to <prefix>[:<exec_prefix>]
> > 'import site' failed; use -v for traceback
> > Python 1.5.2 (#6, Nov 18 1999, 17:19:05)  [GCC 2.8.1] on ms-dos7
> > Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> > 
> > 
> > I've tried everything I can think of to set PYTHONHOME and PYTHONPATH
> > and python ALWAYS starts up this way.
> 
> If I remember (this was long long time ago), I had to remove case
> sensitivity for library files by setting some dummy environment
> variable.

You surely mean the PYTHONCASEOK variable. If that doesn't help, 
simply comment these parts out of the source code... DOS is not going 
to use them anyway.


--Hans Nowak (zephyrfalcon at hvision.nl)
Homepage: http://fly.to/zephyrfalcon
You call me a masterless man. You are wrong. I am my own master.



From gmcm at hypernet.com  Mon Nov  1 12:00:56 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Mon, 1 Nov 1999 12:00:56 -0500
Subject: commas vs. spaces. vs. the legions of new lines
In-Reply-To: <7vj35t$8p1$1@nnrp1.deja.com>
Message-ID: <1270647588-25358408@hypernet.com>

G. Lewis wrote:
> 
> Geez, you peoples, quit sniping on the folks posting their little
> modules lately (if i may snipe at you to say)!
> 
> So the unix old-timers <wink> know a billion and one ways to
> convert Dos to Unix text files.  Why discourage someone from
> making a pure python implementation... 

I will make the observation that one can classify Pythonistas 
into those who think recoding common utilities in pure Python 
is evil, and those who think it a worthy goal, with very few 
people in the middle.

If you live all your life under one OS, I suppose the recoding 
looks like wasted effort. But if you're always switching around, 
or have multiple OSes on your network, having predictable 
pure Python versions is a delight.

(As for converting Unix line endings to DOS - almost all 
Windows editors can handle Unix line endings.)

- Gordon



From gmcm at hypernet.com  Wed Nov 24 19:45:21 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 24 Nov 1999 19:45:21 -0500
Subject: filter function question - list processing in Python
In-Reply-To: <81ho50$g15$1@nnrp1.deja.com>
Message-ID: <1268631776-5935125@hypernet.com>

lobozc at my-deja.com writes:

> 2. I suspect that the second part of my posting may be a valid
> postulate then. Since list processing is so useful Python should
> consider adding a good set of primitives to the core library -
> not as an external package.

A good set of primitives with intuitive interfaces and obvious 
names that raises lots of excitement would probably be 
warmly welcomed.

But, I suspect, those who regularly use filter and reduce are 
vastly outnumbered by those who aren't sure what they do. So 
rabid enthusiasm may be hard to come by.

all-functional-programmers-are-rabid-ly y'rs

- Gordon



From robin at alldunn.com  Mon Nov 22 14:34:25 1999
From: robin at alldunn.com (Robin Dunn)
Date: Mon, 22 Nov 1999 11:34:25 -0800
Subject: Building wxPython [was Re: wxPython and DISLIN work nicely together]
References: <80s0b4$g5n@mail.psy.uva.nl> <19991117090214.B807152@vislab.epa.gov> <017a01bf31f8$6bbea590$1a25d2d1@jenkondev.com> <19991119064944.A996886@vislab.epa.gov> <002301bf32b5$97a5d440$0301a8c0@easystreet.com> <19991119134422.A1004588@vislab.epa.gov <19991122135553.A1255406@vislab.epa.gov>
Message-ID: <01be01bf3520$96f29890$1a25d2d1@jenkondev.com>

> Robin Dunn:
>  |This makes me wonder how you built wxWindows at all if these errors are
in
>  |the wxWindows headers...  What about the wxWindows sample apps, do they
>  |compile?
>
> Beats me but wxWindows worked out of the box.

Well, it makes sense now that you've discovered that Python headers define
NULL and wxWindows never includes Python.h anywhere except in wxPython code.


> With the helpers.cpp NULL hack I just described, it got to the point of
> linking wxcmodule.so, and failed trying to find wxPyHelpers:
>
>     CC  -shared -all wx.o windows.o windows2.o windows3.o events.o misc.o
misc2.o gdi.o mdi.o controls.o controls2.o cmndlgs.o stattool.o frames.o
image.o printfw.o sizers.o
clip_dnd.o -L/home/rhh/software/wxGTK-2.1.11/utils/wxPython/src
`wx-config --libs`  -lwxPyHelpers   -o wxcmodule.so
>     C++ prelinker: warning: could not locate library -lwxPyHelpers;
assuming /usr/lib/libwxPyHelpers.a
>     ld32: FATAL   9  : I/O error (-lwxPyHelpers): No such file or
directory
>     gmake: *** [wxcmodule.so] Error 2
>

libwxPyHelpers.so is the first thing built in the wxPython/src directory.
It hasn't been installed yet until you get to the build.py -i step, but it
does exist in the current directory.  Maybe your linker needs a "-L." ?
Setting OTHERLFLAGS="-L." in build.local might help.


> build.py assumes that the utils/ directory is named wxPython.  I just left
> it wxPython-2.1.11 since the README.txt file said
"preferably...wxPython").
>
> I'm also building under the wxGTK build tree's utils/ dir since I didn't
> see whether wxPython expected to build under the wxGTK build or installed
> trees, and under the build tree made more sense.  Judging from the default
> search path of -L$(WXWIN)/utils/wxPython/src for files in the wxPython
> build tree (WXWIN being the wxGTK installed tree), now I'm changing my
> guess.  Which should it be?
>

By default build.py expects things to be just as they are in the CVS tree,
which is $(WXWIN)/utils/wxPython/src.  You can override this in your
build.local by setting WXPSRCDIR like this:

    WXPSRCDIR = "~/MyStuff/wxPython-2.1.11/src"


I've been updating the build instructions over the last few days to help
people around issues like this and others.  There is version as of a few
days ago at
http://starship.python.net/pipermail/wxpython-users/1999-November/001074.htm
l, and there has also been a few more changes since then.

--
Robin Dunn
Software Craftsman
robin at AllDunn.com
http://AllDunn.com/robin/
http://AllDunn.com/wxPython/  Check it out!





From hinsen at cnrs-orleans.fr  Wed Nov 17 04:39:57 1999
From: hinsen at cnrs-orleans.fr (Konrad Hinsen)
Date: 17 Nov 1999 10:39:57 +0100
Subject: telnetten vanuit een script
References: <Pine.LNX.4.10.9911121516030.28857-100000@semios.c75.net> <19991112201259.A6212@optiplex.palga.uucp>
Message-ID: <m3hfiltq1e.fsf@chinon.cnrs-orleans.fr>

Gerrit Holl <gerrit.holl at pobox.com> writes:

> Telnet moet van een terminal lezen, dus je kan expect gebruiken. Expect is
> een scripting taal voor dit soort dingen; er is echter ook een python module,

Juist, maar er is ook en python module voor telnet (telnetlib), en die maakt
het nog eenvoudiger. Voor meer informatie:

 http://python.via.ecp.fr/doc/current/lib/module-telnetlib.html

-- 
-------------------------------------------------------------------------------
Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.55.69
Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais
-------------------------------------------------------------------------------



From fdrake at acm.org  Fri Nov  5 11:50:39 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Fri, 5 Nov 1999 11:50:39 -0500 (EST)
Subject: Drawing (directed) graphs
In-Reply-To: <000e01bf27ab$5548aba0$3acbd9c2@peridot.optichrome.com>
References: <19991105160912.502.rocketmail@web602.mail.yahoo.com>
	<000e01bf27ab$5548aba0$3acbd9c2@peridot.optichrome.com>
Message-ID: <14371.2783.885109.27130@weyr.cnri.reston.va.us>

Adrian Eyre writes:
 > This is why it annoys me that some of the scripts in the
 > standard Python library are not platform independent, but
 > rely on non-portable features of posix.

  If you know of ways to reduce the platform-dependence of these,
please feel free to submit patches!


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From philh at vision25.demon.co.uk  Mon Nov 29 20:26:17 1999
From: philh at vision25.demon.co.uk (phil hunt)
Date: Tue, 30 Nov 1999 01:26:17 +0000 (GMT)
Subject: Htmlizing text
In-Reply-To: <19991129170705.A314@better.net>
Message-ID: <Pine.LNX.4.05.9911300126100.7546-100000@vision25.demon.co.uk>

On Mon, 29 Nov 1999, William Park wrote:

> On Mon, Nov 29, 1999 at 09:03:46PM +0000, Phil Hunt wrote:
> > 
> > Is there a function in the standard Python library to HTML-ize text,
> > i.e. to replace 'a > b & c < d' with 'a &gt; b &amp; c &lt; d'?
> > 
> > (I realize this can be done with regular expressions, but is there
> > an easy way?)
> > 
> > -- 
> > Phil Hunt - - - phil at comuno.com
> 
> The following function replaces [<&>] unconditionally.
>     def escapeall(s):
> 	s = string.replace(s, '&', '&amp;')		# do this first
> 	s = string.replace(s, '<', '&lt;')
> 	s = string.replace(s, '>', '&gt;')
> 	return s
> I used this in my 'HTMLtag' module which generates HTML tags.
> 
> William.
> 


Thanks.

-- 
***** Phil Hunt ***** send email to phil at comuno.com *****
Moore's Law: hardware speed doubles every 18 months
Gates' Law: software speed halves every 18 months 































From nkauer at my-deja.com  Mon Nov  1 22:11:42 1999
From: nkauer at my-deja.com (nkauer at my-deja.com)
Date: Tue, 02 Nov 1999 03:11:42 GMT
Subject: extension modules and multiple interpreters
Message-ID: <7vlkpc$453$1@nnrp1.deja.com>

I run the following test script on Apache-1.3.6/PyApache-4.19
on Linux 2.2.5 (RedHat 6.0) with only one httpd child and the
module mxDateTime compiled into Python/PyApache, so there is
no dynamic loading of a shared library.

------------------------------------------------------------
print "Content-type: text/plain"
print
print
import sys, traceback
# just checking ...
if not sys.modules.has_key('mxDateTime'):
   print 'No module mxDateTime in sys.modules'
   print

import mxDateTime
print "sys.modules['mxDateTime'] = ", sys.modules['mxDateTime']
print "mxDateTime.__dict__['now'] = ", mxDateTime.__dict__['now']
print
print "calling mxDateTime.now() gives:",
try:
   print mxDateTime.now()
except:
   print
   traceback.print_exc(file=sys.stdout)
--------------------------------------------------------------
First time http://localhost/mytest.py gives
*******************************
No module mxDateTime in sys.modules

sys.modules['mxDateTime'] =  <module 'mxDateTime' (built-in)>
mxDateTime.__dict__['now'] =  <built-in function now>

calling mxDateTime.now() gives: 1999-10-28 18:28:49.75
*******************************
And again http://localhost/mytest.py now gives
*******************************
No module mxDateTime in sys.modules

sys.modules['mxDateTime'] =  <module 'mxDateTime' (built-in)>
mxDateTime.__dict__['now'] =  <built-in function now>

calling mxDateTime.now() gives:
Traceback (innermost last):
  File "/local/web/alpha/docroot/mytest.py", line 16, in ?
    print mxDateTime.now()
TypeError: call of non-function (type None)
*******************************
After that experience, I eliminated all dependencies on module
DateTime/mxDateTime in my Python code, but now I get
the same error whenever something in module MySQLdb is
accessed for the second time ...

The mxDateTime author Marc Lemburg <mal at lemburg.com> writes:
"[...] indicates that PyApache (or perhaps the Python finalizer)
has cleared the module's namespace... which is bad, since
extension modules can typically only initialize themselves *once*."
Marc writes further:
"[T]his is really odd: the 'now' symbol still refers to the
existing function while the module seems to return None as
attribute."

The PyApache coordinator Lele Gaifax <lele at seldati.it> writes:
"I can confirm that the problem lives in the cleanup mechanism.
Python clears up its dictionaries, and doesn't allow a
reinit, as Marc pointed out.  This is not a PyApache fault: you
can get the very same result running the pysrv demo [...].
In fact, every use of such modules mixed to multiple interpreters
should cause the problem."

Apparently, the fact that a module works with regular Python
does not guarantee that it also works with PyApache (or pysrv).

Lele Gaifax writes:
"[...]I cannot see a way out, if not digging in Python's internals.
[D]oes someone know if the problem has been raised before in the
Python community?"

Please, could someone with Python internals expertise explain this
issue?  How can I tell in advance if an extension module will work
with PyApache?  Is it possible to trick Python into cleanly
re-initializing a module?  Is there any hope that I can run my
existing Web site with the Python interpreter embedded into
Apache's httpd?

Nikolas


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



From a.eyre at optichrome.com  Fri Nov  5 08:36:09 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Fri, 5 Nov 1999 13:36:09 -0000
Subject: Drawing (directed) graphs
In-Reply-To: <3823f33e.11772788@news.demon.co.uk>
Message-ID: <000a01bf2792$b83ff150$3acbd9c2@peridot.optichrome.com>

> Good thing there is a platform independent Python graphics library!

How can there be. Some OSs (Minix) don't support graphics at all...

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From jblazi at netsurf.de  Tue Nov 16 12:57:48 1999
From: jblazi at netsurf.de (Janos Blazi)
Date: Tue, 16 Nov 1999 18:57:48 +0100
Subject: Beginning programming with Python
References: <sng0ychgwt.fsf@motorola.com> <382B3A53.5816B2@callware.com> <382B3D79.74CD3E0F@email.sps.mot.com> <80p9ds$ppo$1@newshost.accu.uu.nl> <asteroidal-80psfo/INN-2.2.1/bat@broadway.news.is-europe.net> <38312D55.A36EB56D@iqsoft.hu> <antigorite-80rol2/INN-2.2.1/attentive@broadway.news.is-europe.net> <19991116173142.A31379@optiplex.palga.uucp>
Message-ID: <assonant-80s673/INN-2.2.1/basilar@broadway.news.is-europe.net>

> the less time it costs to program it, the more time it costs to run it,
> the more time it costs to program it, the less time it costs to run it.

Is this some law of nature I missed at university?
But obviously, as far as raw speed is concerned Python cannot compete with
C. Nobody can attempt to deny that. So if speed is an issue for you, use C.
Or even assembly language.
If you do not like Python, use C anyway. Why not? Actually I do not care.

Janos Blazi





From prestonlanders at my-deja.com  Tue Nov 30 16:22:12 1999
From: prestonlanders at my-deja.com (Preston Landers)
Date: Tue, 30 Nov 1999 21:22:12 GMT
Subject: Removing null items
References: <821761$s8u$1@nnrp1.deja.com>
Message-ID: <821f5u$2m5$1@nnrp1.deja.com>

While not especially ingenious, this method does modify the list in
place, and is different.  Well, only in the sense that it is basically
the equivelent inverse of what you did. ;-)

def remove_false_from_list(the_list):
   for item in the_list:
      if not item:
         the_list.remove(item)
   return the_list

---Preston

In article <821761$s8u$1 at nnrp1.deja.com>,
  sragsdale at my-deja.com wrote:
> So I've got a very simple desire: to remove all false (I.E. None,
0, '',
> []) items from a list.
>
> As it is I've got this function which works fine, but it offends my
> sight.  Is there any better (or at least different) way to do this?
>
> #################################################################
> # return non-null items from the first list
> def foo(lst):
>     rlist = []
>     for i in lst:
>         if i:
>             rlist.append(i)
>     return rlist
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
>

--
|| Preston Landers <prestonlanders at my-deja.com> ||


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



From mikael at isy.liu.se  Mon Nov 15 08:49:11 1999
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Mon, 15 Nov 1999 15:49:11 +0200 (MET)
Subject: [Python] header (fwd)
In-Reply-To: <002a01bf2f75$85166620$3acbd9c2@peridot.optichrome.com>
Message-ID: <XFMail.991115154911.mikael@isy.liu.se>

On 15-Nov-99 Adrian Eyre wrote:
 > > Huh? I filter (in Pegasus mail) on "python-list@" in To: and 
 > > CC: and get them all.
 >  
 >  I filter "From:" for "python-list-admin at python.org"
 >  
 >  Does that not work?

Are you sure you don't mean "Sender:" instead of "From:"? I, at least,
get the actual author at "From:", and "python-list-admin at python.org" at
"Sender:".

email-is-magic-ly y'rs

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    15-Nov-99
Time:    15:45:29

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



From xtian_the_great at hotmail.com  Mon Nov  1 19:26:32 1999
From: xtian_the_great at hotmail.com (Xtian Muirhead)
Date: Tue, 2 Nov 1999 13:26:32 +1300
Subject: dynamically generating function assignent
References: <51C9F3C392D0D111A15600A0C99E64A9AF5605@localhost>
Message-ID: <7vlb70$95njn$1@titan.xtra.co.nz>

Bjorn Pettersen <bjorn at roguewave.com> wrote in message
news:51C9F3C392D0D111A15600A0C99E64A9AF5605 at localhost...
> You can do:
>
> fn = eval('func_' + 'a') # look it up in the environment
> fn()
>
> -- bjorn
>
> > -----Original Message-----
> > From: anantk at my-deja.com [mailto:anantk at my-deja.com]
> > Sent: Monday, November 01, 1999 11:19 AM
> > To: python-list at python.org
> > Subject: dynamically generating function assignent
> >
> >
> > Hi,
> >
> >   Here is my problem. I have defined say 12 pairs of functions
> >
> > func_a()
> > parse_func_a()
> >
> > func_b()
> > parse_func_b()
> >  .
> > .
> > func_l()
> > parse_func_l()
> >
> >   A file contains a series of strings which are the names of these
> > functions eg:
> >
> > a
> > c
> > f
> > b
> >
> >   Now for each of these strings I need to call the two functions
> > func_<x>(), and parse_func_<x>(), where <x> is the name of
> > the function
> > that string refers to.
> >
> >   I know you can do something like
> >   call = func_a, and call effectively gets assigned to func_a. My
> > question is how can I generate the RHS of this assignment dynamically?
> > So that all I have to do is
> >
> >   <variable> = <next line in file>
> >   call1 = func_<variable>
> >   call2 = parse_func_<variable>
> >   call1()
> >   call2()
> >
> >   Any help would be appriciated.
> >
> >   Thanks
> >
> > -- Anant

Another way would be to load all the functions into a dictionary.
e.g:

def func_1():
    <blah>

def parse_func_1():
    <blah>

...

functions = {'1':(func_1,parse_func_1),
             'a':(func_a, parse_func_a),
                ...}

Then you could look up call1 and call2:

call1, call2 = functions[valueFromFile]
call1()
call2()

This could be easier to catch cases where the value from the file
doesn't match any of the functions defined. (And I prefer this kind
of static lookup to a dynamic lookup. That's just a bias, I guess.
Suffice it to say that both of these ways are nicer than a whole heap
of elifs. (I guess if you had hundreds of these functions, you wouldn't
want to have to type out all of them in a dictionary (with the
attendant likelihood of making an error).))

Xtian

xtian at regurgitator.zzn.com

"Most people don't realise that large pieces of coral, which have been
painted brown and attached to the skull by common wood screws, can make
a child look like a deer."






From sekter at mail.matav.hu  Mon Nov 15 08:10:05 1999
From: sekter at mail.matav.hu (Arpad Kiss)
Date: Mon, 15 Nov 1999 14:10:05 +0100
Subject: PIL 1.0 on win32?
References: <009b01bf2f5e$fef3b900$029f38c3@geometria.hu> <01ee01bf2f61$96097a80$f29b12c2@secret.pythonware.com>
Message-ID: <004701bf2f6a$bd67f990$029f38c3@geometria.hu>

Fredrik Lundh wrote:
> Arpad Kiss wrote:
> > I have a PIL based Zope product I have been developing on a Linux box. I
> > just wanted to test it on a Windows NT(sp5), but it doesn't work.:-(
> > At first it didn't recognize my fonts, because the readline method on
win32
> > returns 10+13 at the end of the lines and in ImageFont.py only \n is
> > checked.
>
> PIL opens the font files in binary mode, so this
> shouldn't be a problem.  you probably messed
> something up when you transferred the font
> files to the PC...

I have created a .tar.gz then extracted it with WinZip.

>
> > After I  solved it I got an error message(memory error in line 151 of
> > ImageDraw.py) when I wanted to draw some text.
> > When I commented out the text drawing code in my program, then I got
back a
> > picture, but it is mirrored!
>
> this sounds really weird -- but at least the
> memory error might be due to bogus font data...

Ok. I will try to replace my fonts.

>
> what win32 build do you use?  can you post some
> code?

Python1.5.2 and pil-win32-991101.zip
I will write a minimal example.

>
> > Of course on linux box my program works fine.
> > I just want know who else use PIL 1.0 on win32 and what experiences they
> > have. Should this text drawing function work? Is it my fault?
>
> fwiw, PIL is being developed on win32...

Maybe a minimal example(without Zope) can help me, if cannot then I'll post
it.
Thanks,
Arpad





From anantk at my-deja.com  Mon Nov  1 13:19:02 1999
From: anantk at my-deja.com (anantk at my-deja.com)
Date: Mon, 01 Nov 1999 18:19:02 GMT
Subject: dynamically generating function assignent
Message-ID: <7vklii$crc$1@nnrp1.deja.com>

Hi,

  Here is my problem. I have defined say 12 pairs of functions

func_a()
parse_func_a()

func_b()
parse_func_b()
.
.
func_l()
parse_func_l()

  A file contains a series of strings which are the names of these
functions eg:

a
c
f
b

  Now for each of these strings I need to call the two functions
func_<x>(), and parse_func_<x>(), where <x> is the name of the function
that string refers to.

  I know you can do something like
  call = func_a, and call effectively gets assigned to func_a. My
question is how can I generate the RHS of this assignment dynamically?
So that all I have to do is

  <variable> = <next line in file>
  call1 = func_<variable>
  call2 = parse_func_<variable>
  call1()
  call2()

  Any help would be appriciated.

  Thanks

-- Anant
   anantk at yahoo.com


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



From fredrik at pythonware.com  Thu Nov 18 05:33:03 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 18 Nov 1999 11:33:03 +0100
Subject: __str__ vs. __repr__
References: <000001bf275c$02370d20$1d2d153f@tim> <7vvni7$19q0$1@nntp6.u.washington.edu> <382BEC79.EC65913D@maxtal.com.au> <382C6430.8DC07E3E@lemburg.com> <01fd01bf2dce$d2c03cf0$f29b12c2@secret.pythonware.com> <942499879snz@vision25.demon.co.uk> <m3n1shp2uu.fsf@atrus.jesus.cam.ac.uk> <942717443snz@vision25.demon.co.uk>
Message-ID: <012f01bf31b0$50570ab0$f29b12c2@secret.pythonware.com>

Phil Hunt <philh at vision25.demon.co.uk> wrote:
> How about:
> 
>    x = a + b + c
>    x = (a + b) + c
>    x=a+b+c
> 
> I'd guess these all produce the same bytecode.

footnote: the abc80 computer I mentioned
in my earlier post had an extra byte code
for this purpose.  users can buy that you
add spaces and reformat their code, but
they won't accept that you remove their
parentheses...

> I'd also not be surprised if you could consistently change the
> name of a local variable within a function without changing bytecodes,
> i.e. Python internally doesn't know the names of local variables.

print myfunction.func_code.co_names

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
(news: just posted an errata for the 'first printing') -->




From fredrik at pythonware.com  Sat Nov 13 06:55:17 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 13 Nov 1999 12:55:17 +0100
Subject: Random text generator
References: <199911120349.OAA01177@envy.fulcrum.com.au> <y0jk8nntzvn.fsf@vier.idi.ntnu.no>
Message-ID: <01e101bf2dcd$f4bb6290$f29b12c2@secret.pythonware.com>

Magnus L. Hetland <mlh at vier.idi.ntnu.no> wrote:
> >    The function you want is whrandom.choice(). See
> 
> Or - rather, random.choice() (which uses whrandom.choice, if
> available)... No?

it would probably be a bit clearer (at least to newcomers)
if "random" is was the official interface, and whrandom was
treated as an implementation module.  but it doesn't look
that way in the current version of the library reference...

(and the eff-bot guide is obviously wrong on this one. not
sure I'll add that to the errata, though ;-)

</F>




From kuncej at mail.conservation.state.mo.us  Fri Nov  5 11:02:13 1999
From: kuncej at mail.conservation.state.mo.us (Jeffrey Kunce)
Date: Fri, 05 Nov 1999 10:02:13 -0600
Subject: Python Conference Hotel Change
Message-ID: <s822ab8a.094@mail.conservation.state.mo.us>

GVR>...a 15 minute walk across the bridge to historic Georgetown ...

AR>... but isn't it an even shorter stroll from there to Bardo ...

What a hearty lot, these Pythoneers! Others might find these
mid-winter perambulations a bit chilly. :-)

  --Jeff





From morse at harborcom.net  Tue Nov  9 08:03:18 1999
From: morse at harborcom.net (Kevin Dahlhausen)
Date: Tue, 09 Nov 1999 13:03:18 GMT
Subject: Stackless python progress?
References: <38274626.5D8C1EC2@ptway.com>
Message-ID: <38281b64.88310413@news.oh.verio.com>

I must confess I don't know this, but what is the advantage of stackless Python?
Are you referring to the byte-code interpreter?

Brian Haskin <haskin at ptway.com> wrote:

>Just wondering if the work on stackless python is still progressing and
>if  it looks like it will be integrated into the regular distribution in
>the foreseeable future? also has there been any work on combining
>'microthreading' and the stackless stuff?
>
>Brian Haskin




From jfarrell at mincom.com  Wed Nov  3 18:37:31 1999
From: jfarrell at mincom.com (John Farrell)
Date: Thu, 04 Nov 1999 09:37:31 +1000
Subject: How does rexec work?
Message-ID: <3820C73B.9278E532@mincom.com>

I'm confused about rexec and how it works. Consider the following code:
---
import rexec, traceback

renv1 = rexec.RExec()
renv1.r_exec('import string')
print renv1.r_eval("string.upper('This is a test')")

try:
    renv2 = rexec.RExec()
    renv2.r_import('string')
    print renv2.r_eval("string.upper('This is a test')")
except:
    traceback.print_exc()
----
The things I don't understand are:

 (a) In the renv1 example, where does the name string get preserved
to be available in the eval? The relevant code from rexec is included
below, but I don't see how the self.modules dict gets into the
__main__ namespace.
 (b) Why doesn't r_import work like r_exec('import ...')? What's it
supposed to do then?

I guess part of my problem is that I am not 100% familiar with
the Python namespace fiddling stuff yet. Thanks for any help!

John
----
    # Add a module -- return an existing module or create one

    def add_module(self, mname):
        if self.modules.has_key(mname):
            return self.modules[mname]
        self.modules[mname] = m = self.hooks.new_module(mname)
        m.__builtins__ = self.modules['__builtin__']
        return m

    # The r* methods are public interfaces

    def r_exec(self, code):
        m = self.add_module('__main__')
        exec code in m.__dict__

    def r_eval(self, code):
        m = self.add_module('__main__')
        return eval(code, m.__dict__)
----



From evan at tokenexchange.com  Sat Nov  6 00:04:01 1999
From: evan at tokenexchange.com (Evan Simpson)
Date: Fri, 5 Nov 1999 23:04:01 -0600
Subject: making the Python case to my dept.
References: <9911051740130D.02530@william.jinsa.org>
Message-ID: <7GOU3.272$Ae3.194028@news-dal.corridex.com>

William Ahern <wahern at jinsa.org> wrote in message
news:9911051740130D.02530 at william.jinsa.org...
> I have never used Python, but I'm currently attempting to persuade a
department
> at my university (Geroge Washington) to use Python as an intro language. I
have
> read that it was good for this. From what I hear, the current options on
the
> table are Delphi and Java. I don't know much about Delphi, but I doubt
Java is
> any good for intro classes.

Well, I don't know much about Java, but I've been a professional Delphi
programmer for about five years now.  Up until a year and a half ago, I'd
have said that Delphi would be fine for your purposes.  Then I learned
Python :-)

Now I'm preparing to introduce my children to programming through Python.
I'm starting with their fascination with Alice, the 3-D animation workshop
which even my 6 year old enjoys playing with (though he doesn't script it...
yet).

Cheers,

Evan @ 4-am





From dirkl at home.com  Mon Nov  8 04:32:40 1999
From: dirkl at home.com (Dirk Leas)
Date: Mon, 8 Nov 1999 02:32:40 -0700
Subject: Production Web Deployment?
In-Reply-To: <199911080604.BAA22580@python.org>
Message-ID: <NBBBKMMANMGEEBHPJJECGEGIFJAA.dirkl@home.com>

I've read numerous threads regarding how to deploy Python on the web and
there appears to be no real consensus on what actually performs/scales well
(about 3M hits/month). Is anybody taking lots of hits with httpdapy,
PyApache, ASP+Python, fastcgi, or Zope? I've got a customer who's
development staff is struggling with the complexity of Java Servlets and
they're already using a large 3rd party, vertical Python application (and
several folks already use Python). Peliminary testing showed plain CGI was
just too slow...

Thanks,
D




From nemeth at iqsoft.hu  Thu Nov  4 04:21:20 1999
From: nemeth at iqsoft.hu (Németh Miklós)
Date: Thu, 04 Nov 1999 10:21:20 +0100
Subject: Extending Python dynamically
Message-ID: <38215010.DCFCDBFA@iqsoft.hu>

Dear Pythoners,

I have downloaded DCOracle extension. I have RH6.0, and (at least until
now) I have not installed the source distribution of Python. I have
installed only the binary RPMs. I unpacked DCOracle in a directory and
managed to build the so files (Buffer.so and oci_.so).
How to install then this DCOracle package? I have been trying to figure
out but no success.
I have read ext.pdf (Extending Python with C or C++). As far as I
understood (by reading chapter 2), I have to recompile my Python
interpreter, and have to relink the interpreter against this DCOracle
shared objects.
Does Python not support a more dynamic extension mechanism (based on
dlopen())? If I want to add an extension in C, shall I always
recompile/relink my Python interpreter? Perl (via XS) and Apache (via
mod_so) supports a kind of dynamic loading mechanism I'd like to have.
Can anyone help me how to get this DCOracle work on my system?

Miklos Nemeth
IQSOFT

PS: I am quite new to Python, but an experienced OO (C++, Java)
programmer. I was amazed by Python as a language. Now I am evaluating
Python to determine the quality/applicability of the facilities
necessary for my daily works (such as Oracle7 database access, CORBA
(over SSL is preferred) support, extendability, GUI support, etc.),
which mainly are mission critical business applications.
I have also started evaluating Zope. After the first two hours I found
Zope is a fantastic product, I am curious about how people did dare
implement a so complex application (ie. an application server) in an
interpreted "scripting language". These type of software pieces are
implemented in C (Apache, AOLserver), C++ (BroadVision). So I decided I
learn Python and assemble a development environment for Python
programming.

NM





From fredrik at pythonware.com  Mon Nov  8 17:48:04 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 8 Nov 1999 23:48:04 +0100
Subject: Image buttons with Tkinter
References: <38270596.A391995C@parlant.com>
Message-ID: <01fd01bf2a3b$52cbd3b0$f29b12c2@secret.pythonware.com>

Thomas Lane wrote:
> I'm trying to create a button that displays an image instead of text,
> but I don't seem to be having any luck. The button just comes up gray.

see:

    http://www.python.org/doc/FAQ.html#4.69
    4.69. Images on Tk-Buttons don't work in Py15?

</F>

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




From marco.musazzi at sofositalia.it  Fri Nov  5 15:48:57 1999
From: marco.musazzi at sofositalia.it (Marco Musazzi)
Date: Fri, 5 Nov 1999 21:48:57 +0100
Subject: Template for CGI
References: <Pine.LNX.4.20.9911030821220.30064-100000@fep132.fep.ru>
Message-ID: <7vvfpl$o2a$1@pluto.sm.dsi.unimi.it>

>On Tue, 2 Nov 1999, Glenn Kidd wrote:
>> I am a newbie to Python and I have a question.  I want a CGI script that
>> will read a file template and replace certain words with variables.  Ii
>> tried putting the variable names in the file but it is read as palin
text.
>> Is there a way to embed variables into a teplate or a replace function in
>> Python?  Thank you for any help you give.


You can use something like this:

template file:
your template text  your template text %variable% your template text
%variable2%

%xxx% is the placeholder for "variable" xxx (you can replace % with what you
like)

in your python code:

import string

template = open("yourtemplatefile")
templateText = template.read()
template.close()

templateText = string.replace(templateText, "%variable%",
str(value_of_variable))
templateText = string.replace(templateText, "%variable2%",
str(value_of_variable2))
# an so on....


# printing a minimal http header
print "Content-type: text/html"
print

# printing the template filled with the variables
print templateText


You can do it better using a value-filled dictionary:

val = {"variable":value1, "variable2":value2,  etc...... } # you may read
this dict from a shelv db for example...

for k in val.keys():
    templateText = string.replace(templateText, "%"+ k + "%", val[k] )

# printing a minimal http header
print "Content-type: text/html"
print

# printing the template filled with the variables
print templateText

And you can use both the mothods at the same time...

happy coding
Marco






From neelk at brick.cswv.com  Tue Nov  9 20:17:08 1999
From: neelk at brick.cswv.com (Neel Krishnaswami)
Date: 10 Nov 1999 01:17:08 GMT
Subject: __str__ vs. __repr__
References: <000401bf25b9$88ea57e0$2e2d153f@tim>  <199911040157.UAA11908@eric.cnri.reston.va.us>  <046101bf26d4$2dd1a720$01ffffc0@worldnet.att.net>
Message-ID: <slrn82hhtv.ppi.neelk@brick.cswv.com>

Emile van Sebille <emile at fenx.com> wrote:
>
>It seems to me that issues arise when using the interactive interpreter.
>I suspect that little useful work is done in this mode, and is used
>predominately for quick testing while creating python programs.

I'd like to politely disagree . Combined with Tim Peters and Barry
Warsaw's emacs py-mode, Python's interactive mode is the single most
useful feature of the language for me. With it, I can test each and
every method, loop, and statement as I code. This does wonders for my
understanding of my code. And not incidentally, reliability goes way
up as well.

I learned Perl before Python, but my Python skill surpassed my Perl
skill after a much smaller time investment. I lay this pretty much
wholely at the feet of the interactive interpreter -- I could find out
the answers to puzzles without a tedious edit-save-run cycle.


Neel



From ionel at psy.uva.nl  Tue Nov  2 17:32:29 1999
From: ionel at psy.uva.nl (Ionel Simionescu)
Date: Tue, 2 Nov 1999 23:32:29 +0100
Subject: Drawing (directed) graphs
References: <ye1bt9das2p.fsf@ex.ac.uk>
Message-ID: <7vnop8$bqt@mail.psy.uva.nl>

R.M.Everson <R.M.Everson at ex.ac.uk> wrote in message
news:ye1bt9das2p.fsf at ex.ac.uk...

| Does anyone know of Python code for drawing graphs of the
| graph-theoretic type? Ideally, I'd like to be able to interactively
| add nodes and edges etc and add data to structures associated with the
| nodes and edges.
|

The closest library I know is OGL - Object Graphics Library, which comes
with wxPython 2.1.5. ( http://alldunn.com/wxPython/ ).

You can draw  graphical objects of various shapes and connect them with
lines or splines.
You can derive new objects easily, define connection points, nest the
shapes, resize objects.

A problem is that you need to do some (maybe, much) of the layout yourself.

You get some support from composite shapes which are able to manage a given
layout, but there's still a lot of work to do for getting close to something
like graphviz.

ionel


------
extract from:
http://web.ukonline.co.uk/julian.smart/wxwin/contrib/ogl/ogl.htm

Features of OGL
A range of predefined shape classes, such as wxRectangleShape,
wxPolygonShape, wxEllipseShape.
Multiline and spline wxLineShape connecting node shapes: the arcs move with
the nodes.
Arcs join neatly at the perimeter of shapes.
Arcs can either join at attachment points or as if connected to the centre
of the node.
Resizing handles.
Multiple text regions for complex shapes.
Divided rectangles: rectangles with an arbitrary number of vertical
divisions.
Composite shapes.
Simple 'metafile' shapes which can be resized and rotated.
Bitmap shapes.
wxDiagram class manages a list of shapes, and implements input/output.

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

The relationship between OGL and Hardy
OGL is the core graphics code for Hardy, a meta-CASE tool written with
wxWindows.
There are various features that Hardy implements over and above the OGL
functionality, including a symbol library, node and arc symbol editors,
attachment point editor, junction node functionality, and various editing
facilities. It is possible that contributors could extend OGL to encompass
these areas in future.







From nikolai.kirsebom at siemens.no  Fri Nov  5 04:58:58 1999
From: nikolai.kirsebom at siemens.no (Nikolai Kirsebom)
Date: Fri, 05 Nov 1999 09:58:58 GMT
Subject: Accessing Word object model from Python COM server
Message-ID: <3822a7d2.251040356@news.mch.sni.de>

I'm using a Python COM server when testing a VBA (Word97) system.  I
would like to be able to access the Word object model from the methods
in the COM server.  Does anyone know how this can be done.

I would like the following setup:

VBA-code:
** Initialization:
Dim Srv As Object
Set Srv = CreateObject("PythonCOMserver")
Srv.SetRef Application

** Action handling:
Dim doc As Word.Document
doc = Srv.GetNew


Python COM server code:
def SetRef(self, applObj):
  self.ApplObj = appObj
  return 0

def GetNew(self):
  doc = self.AppObj.Documents.Add()
  # some additional handlnig
  return doc


When the GetNew method in the Python COM server is executed, an error
indicates that the name 'Documents' is invalid.  The self.ApplObj is
of type PyIDispatch (I think), however may it must be an 'instance' in
order to be able to access the object model.  If I instead of
accessing the object model just returns the object (self.AppObj) and
makes use of the returned object back in the VBA code, I can use the
object model - so the round-trip seems to work.

Thanks for any help/hints.

Nikolai Kirsebom






From jchan at cohocanoes.com  Sat Nov  6 02:54:24 1999
From: jchan at cohocanoes.com (jchan at cohocanoes.com)
Date: Sat, 06 Nov 1999 07:54:24 GMT
Subject: Zope comparisons? (ASP, PHP, etc)
References: <3821C5F6.84543AF2@yahoo.com> <06lU3.259$Ae3.186544@news-dal.corridex.com>
Message-ID: <Q8RU3.32341$it.679981@news2.rdc1.on.home.com>

Wow...this is by far one of the best explainations of what Zope is capable 
of doing. Wonderful description, Evan. Perhaps your prose should be        
incorporated into Zope's site.

Evan Simpson <evan at tokenexchange.com> wrote:
> There has been a lot of discussion lately on the Zope mailing list
> (zope at zope.org) about making this sort of information more readily available
> on the website, but here's a brief intro:

> Zope has several parts, which work well together, but are all optional:

> 1.  ZServer: A Medusa-based multi-threaded asynchronous server which accepts
> HTTP, FTP, and monitor (special Python client) connections.  Usually, it
> passes requests from these connections to...

> 2. ZPublisher:  An ORB which converts HTTP, FTP, XML-RPC, WebDAV, etc.
> requests into traversal of an object hierarchy and returns the result of
> rendering the target object.  These objects are normally stored in...

> 3. ZODB: A versioning, transaction-based object store with automatic object
> persistence and conflict handling.  ZODB stores an object hierarchy as a
> single file (2GB limit being worked on) and appends new versions of objects
> when their request transaction is comitted.  This allows extensive undo, and
> requires occasional packing to eliminate old, unwanted versions.  Most HTML
> is generated by standard Zope objects using...

> 4. DTML: An embedded language similar to ASP and PHP, except that
> expressions are written in Python and have access to the full power of the
> rest of Zope, including...

> 5. Acquisition: A feature of standard Zope objects which allows them to
> change their behavior automatically based on the context in which they are
> called.  Hard to describe, but very powerful, as are...

> 6. Products: New objects or collections of objects can be added to Zope by
> simply installing a Python package in the Products subdirectory and
> resetting the server.  Current Products include database adapters, threaded
> discussions, virtual host managers, and many others.  Of course, if you
> don't want to learn Python you can always create...

> 7. ZClasses: Object classes defined through the web management interface by
> assembling standard Zope parts.  These can provide a wide range of
> customized objects.

> I haven't really gotten into the user/role/permission security model, or the
> extensive web management interface through which you typically interact with
> Zope.  I may have forgotten other bits, too.  There's a lot there :-)







From Markus_Strickler at everest-design.de  Thu Nov 11 04:02:01 1999
From: Markus_Strickler at everest-design.de (Markus Strickler)
Date: Thu, 11 Nov 1999 10:02:01 +0100
Subject: Python 1.5.2 on NeXTStep
Message-ID: <002501bf2c23$6aceabb0$2d131bc3@everestdesign.com>

Hi!
Has anyone managed to compile Python 1.5.2 on NeXTStep 3.3? I'm currently
running an older Python version and would like to upgrade. However 1.5.2
won't compile. Can anyone give me start what config parameters I need.

Thanks,

Markus




From fredrik at pythonware.com  Tue Nov  2 03:53:17 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 2 Nov 1999 09:53:17 +0100
Subject: Tkinter canvas blow-up:  bbox(ALL) == None
References: <19991101102103.A134035786@vislab.epa.gov>
Message-ID: <013c01bf2510$57457f20$f29b12c2@secret.pythonware.com>

Randall Hopper <aa8vb at yahoo.com> wrote:
>      Apparently if the bounds of the items in a Tk canvas widget exceed
> MAXINT, bbox(ALL) returns None.  The attached Python script demonstrates
> this.
> 
>      Intuitively it seems like this is a bug since the contents of the
> canvas are maintained in floating point; integer bounds shouldn't be
> involved, should they?
> 
>      Is this a Tk bug or a Tkinter bug?

both.

Tkinter uses _getints instead of _getdoubles to
convert the bounding box to a tuple...

...but the reason you get None instead of an over-
flow error is that Tk returns an empty string in this
case (at least in 8.0.5).

</F>

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




From gerglery at usa.net  Fri Nov 19 04:56:11 1999
From: gerglery at usa.net (Alternative Fluffy)
Date: 19 Nov 1999 04:56:11 -0500
Subject: Beginning programming with Python
References: <Pine.LNX.4.10.9911191043080.2313-100000@essle.valdyas.org>
Message-ID: <877ljest38.fsf@slip-32-100-244-40.ma.us.prserv.net>

Boudewijn Rempt <boud at rempt.xs4all.nl> wrote:

> You needn't guess about that, I guess ;-). I didn't know there
> was a design team for Python, I though it was Guido's invention.
> But I still wonder about the double == - if I can't assign in
> an if, why insist on ==?

Here is one example of why they need to be different:

>>> y = a = 0
>>> z = a == 0
>>> print y
0
>>> print z
1




From phd at phd.russ.ru  Thu Nov 25 06:51:26 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Thu, 25 Nov 1999 11:51:26 +0000 (GMT)
Subject: threading and timeouts
In-Reply-To: <943504970.076.346131773@cs.usyd.edu.au>
Message-ID: <Pine.LNX.4.21.9911251150530.11528-100000@fep132.fep.ru>

On Thu, 25 Nov 1999, Piers Lauder wrote:
> The only option I can think of is to re-write the smtplib module to use
> `select', but does anyone know of a simpler solution?

   Fork off a child that will send the mail.

Oleg.
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From greg.ewing at compaq.com  Mon Nov  8 08:46:28 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Tue, 09 Nov 1999 02:46:28 +1300
Subject: Some more Ruby observations (Re: Comparison between Python and 
 "Ruby")
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp> <7vjtrc$hmm$1@newshost.accu.uu.nl>  <3821A860.5D32AFCC@compaq.com> <slrn824eo2.lsi.neelk@brick.cswv.com>
Message-ID: <3826D434.C083F6FD@compaq.com>

Neel Krishnaswami wrote:
> 
> and since continuation-passing forms are just a hair away from
> bytecode, I am a bit surprised by this news.

I don't see how continuation-passing and bytecode have anything
to do with each other.

> Are there any syntax-tree munging phases to do stuff like common
> subexpression elimination and loop hoisting?

I haven't seen any so far, although I haven't really looked
for these things.

I get the impression that speed isn't a major design goal in
this implementation of Ruby, at least not as much as it is
in CPython. All of Ruby's built-in operations go through a
method lookup, for example, even adding two integers.

Greg



From s323140 at student.uq.edu.au  Tue Nov 16 03:39:01 1999
From: s323140 at student.uq.edu.au (Rob Hodges)
Date: 16 Nov 1999 18:39:01 +1000
Subject: list.without()?
References: <y0jln7zi9pb.fsf@vier.idi.ntnu.no>
Message-ID: <m3d7tazv8a.fsf@phaedrus.zen.xom>

mlh at vier.idi.ntnu.no (Magnus L. Hetland) writes:

> Is there a function or method in the standard language that can
> non-destructively return a list without a specified element? Would it
> be possible to add it as a method to lists?
> 
> i.e:
> 
> # pseudocode
> 
> class list:
>     def without(self,element):
>         temp = self[:]
>         temp.remove(element)
>         return temp

Use a lambda with filter, like this:

>>> a = [1,2,3,5,2,1,4,9,1,3]
>>> a
[1, 2, 3, 5, 2, 1, 4, 9, 1, 3]
>>> b = filter(lambda(elt): elt != 1, a)
>>> b
[2, 3, 5, 2, 4, 9, 3]
>>> a
[1, 2, 3, 5, 2, 1, 4, 9, 1, 3]
>>> 

It would be kinda nice to have such a thing as a list method, now that
you mention it, but this is a pretty reasonable one-liner solution.

-Rob




From tbryan at python.net  Sat Nov  6 03:27:10 1999
From: tbryan at python.net (Thomas A. Bryan)
Date: Sat, 06 Nov 1999 03:27:10 -0500
Subject: __str__ vs. __repr__
References: <000001bf275c$02370d20$1d2d153f@tim> <7vvni7$19q0$1@nntp6.u.washington.edu>
Message-ID: <3823E65E.D9E49CC6@python.net>

Donn Cave wrote:

> repr is used
> in lots of places, like the interactive interpreter,

I think that was Tim's point.  Earlier he said:
"""For that matter, why is repr implied at an interactive prompt given 
a raw expression?  Seems that friendly strings would be more 
appropriate there."""

He wants str to be used at the interpreter since it's supposed to be 
the human readable form.  One of Guido's complaints was that then 
1 and '1' then look the same.

> On the contrary, it's str() that's intended for the machine, and 
> repr() for the human.

>From the library reference:
"""repr (object) 
    Return a string containing a printable representation of an object. 
    This is the same value yielded by conversions (reverse quotes). It 
    is sometimes useful to be able to access this operation as an 
    ordinary function. For many types, this function makes an attempt 
    to return a string that would yield an object with the same value 
    when passed to eval(). 
str (object) 
    Return a string containing a nicely printable representation 
    of an object. For strings, this returns the string itself. 
    The difference with repr(object) is that str(object) does not 
    always attempt to return a string that is acceptable to eval(); 
    its goal is to return a printable string. """

The problem is that this thread is the first time many of us have 
ever thought about the problem.  I've never noticed those two lines
in the library reference, and I haven't seen the str/repr distinction 
clearly made anywhere else.  Furthermore, since the interactive i
nterpreter *does* use repr(), many of us make __repr__ return the 
human readable form so that we can easily test our classes.

So, what I'm hearing is:
1. __str__ *should* be a human readable string
2. __repr__ *should* be a representation that can be eval'd back to the 
   original object
3. This convention is not widely known, but
3b. the interactive interpreter discourages the convention by using 
   repr to show the result of an expression

Therefore, we could switch the convention for str and repr, which breaks
things that are currently following the convention.

Or we could switch the interactive interpreter to use str to encourage 
class writers to follow the convention.  Of course, any class writers 
who intentionally switched __repr__ and __str__ to get nice interactive 
behavior will have to switch the mehtods back.

Since Tim voted for the second option, that's probably what Guido will 
decide. ;)  Either way, Guido still has to deal with non-printing 
characters, string-vs-number distinctions, and other insanity in the 
interpreter. 

who-uses-the-interpreter-anyway-<wink>-ly yours
---Tom



From tbryan at python.net  Tue Nov 30 20:03:35 1999
From: tbryan at python.net (Thomas A. Bryan)
Date: Tue, 30 Nov 1999 20:03:35 -0500
Subject: Python at Source Forge?
Message-ID: <384473E7.5B91C674@python.net>

Has anyone looked at www.sourceforge.com yet?

Perhaps there could be a Parnassus mirror there?
I notice that there is an entry for Perl
http://www.sourceforge.com/project/index.php?form_grp=320
that appears to be a CPAN mirror.

In any case, Python developers of Open Source programs might 
find the resource useful.  

---Tom



From endgamer71 at hotmail.com  Tue Nov 16 06:56:28 1999
From: endgamer71 at hotmail.com (Endgamer)
Date: Tue, 16 Nov 1999 11:56:28 GMT
Subject: Alright, I give up...chalk this one up to experience...
Message-ID: <383145c0.4445469@212.225.36.99>

Alright, I finally give up. I've been bashing at this bloody idea all
week...maybe I'm incredibly stupid (probably), but I can't see why,
when I run this as a script from Windows 95, it doesn't return the
current date setting on my machine. Could somebody put me out of my
misery? I promise I'll stop bothering you <g>.

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

# Date setting program (chunk 1)
import os

olddate=os.popen("date",'r')

olddate=readdate

print olddate

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


Endgamer
Spider Warrior

(Endgamer71 at spamtrap.hotmail.com)
#####################################
Not a quote, just an on-the-spot observation:
You know you're a geek when your PC costs exactly
15 times as much as your stereo. And gives better
sound quality.<sigh>
                              Endgamer
#####################################



From andrew.markebo at telelogic.se  Mon Nov  1 02:43:01 1999
From: andrew.markebo at telelogic.se (Andrew Markebo)
Date: 01 Nov 1999 08:43:01 +0100
Subject: Listbox.. ACTIVE-item is the one selected previously..
References: <ksemebjifu.fsf@telelogic.se> <7vj2ic$ich$1@vvs.superst.iae.nl>
Message-ID: <ksk8o2odx6.fsf@telelogic.se>

The only difference when using anchor is that I get " abcd", it seems
to work fine on Unix, but at home, on my Win98 and WinNT-machines I
get this delayed 'report'.

Smells like a bug.. 

        /andy


  -----------== Posted via Newsfeeds.Com, Uncensored Usenet News ==----------
   http://www.newsfeeds.com       The Largest Usenet Servers in the World!
------== Over 73,000 Newsgroups - Including  Dedicated  Binaries Servers ==-----



From che at debian.org  Fri Nov 12 18:58:17 1999
From: che at debian.org (Ben Gertzfield)
Date: 12 Nov 1999 15:58:17 -0800
Subject: __doc__ print \012 on Debian Gnulinux2
References: <Pine.A41.4.05.9911121848040.37340-100000@maia.sca.uqam.ca>
Message-ID: <ytt66z7xnxi.fsf@gilgamesh.cse.ucsc.edu>

>>>>> "Jean-Bernard" == Jean-Bernard ADDOR <jbaddor at sca.uqam.ca> writes:

    Jean-Bernard> Hey !  I try to make some documentation strings in
    Jean-Bernard> my code but then I use them the result is very
    Jean-Bernard> difficult to read:

*snip*

    Jean-Bernard> How could I make it print return instead of \012?

Try:

print class.method.__doc__

instead of just 'class.method.__doc__'. :)

-- 
Brought to you by the letters A and L and the number 11.
"Wuzzle means to mix."
Debian GNU/Linux maintainer of Gimp and GTK+ -- http://www.debian.org/



From tomh at po.crl.go.jp  Tue Nov 30 21:56:54 1999
From: tomh at po.crl.go.jp (Tom Holroyd)
Date: Wed, 1 Dec 1999 11:56:54 +0900
Subject: Looking for sshlib.py
In-Reply-To: <tBX04.11350$VA6.54975@news-server.bigpond.net.au>
References: <iwI04.9621$VA6.47863@news-server.bigpond.net.au> <19991130080326.A1376@quark.emich.edu> <tBX04.11350$VA6.54975@news-server.bigpond.net.au>
Message-ID: <Pine.LNX.4.20.9912011148280.1206-100000@marvin.crl.go.jp>

Check out SRPSocket, which creates SRP authenticated sockets.

http://www.aist.go.jp/NIBH/~tomh/SRPSocket.tar.gz

The home page for SRP is http://srp.stanford.edu/srp/

SRP stands for Secure Remote Passwords and is a drop in replacement for
Unix passwords that is safe to use over the network.  The main SRP site
has code written in C for telnet and FTP; SRPSocket is pure Python (still
pretty fast though).  The only caveats are that the Python random number
generator isn't really good enough -- you should use a different one, and
the default prime field is 128 bits -- I have a 1024 bit one, and you can
of course generate your own big prime numbers...  (SRP is a variant of
Diffie-Hellman, but better.)

SRP (and SRPSocket) are free from export restrictions as this is only the
authentication stage.  A side effect of the authentication is that you get
a secure, shared key (the length is the output size of the hash used).  
You can use this key to bootstrap a longer key (such as an interleaved SHA
key) or use it to encrypt the socket directly, or whatever.

On Tue, 30 Nov 1999, John Leach wrote:

> I'm looking to write a Python frontend for character-based applications
> accessed via the internet.
> telnet works well but I'd like something with more security (telnet sends
> its password unencrypted I believe).
> I'll try calling ssh from within Python and see if that helps.
> Thanks for your reply,
> John Leach
> http://www.bosware.com.au

Dr. Tom Holroyd
I would dance and be merry,
Life would be a ding-a-derry,
If I only had a brain.
	-- The Scarecrow




From just at letterror.com  Fri Nov  5 06:44:52 1999
From: just at letterror.com (Just van Rossum)
Date: Fri, 5 Nov 1999 12:44:52 +0100
Subject: __str__ vs. __repr__
In-Reply-To: <000001bf275c$02370d20$1d2d153f@tim>
References: <199911041325.IAA14001@eric.cnri.reston.va.us>
Message-ID: <l03102801b44870c20480@[193.78.237.174]>

At 2:04 AM -0500 11/5/99, Tim Peters wrote:
>rendering Python useless to grownups.  How does Python's Unicode story
>relate to this?  Displaying a Unicode string as a pile of \uxxxx (whatever)
>escapes defeats the whole purpose of Unicode.  Ditto displaying it in UTF-7.
>OTOH, for some number of years to come it will be a rare display device that
>*won't* treat Unicode 16-bit values (whether or not UTF-8 encoded) as a pile
>of 8-bit control codes.
[ ... ]
>Until Unicode takes over the world, there's nothing you can do other than
>tell users the truth:  "most of" printable 7-bit ASCII displays the same way
>across the world, but outside of that *all* bets are off.  It will vary by
>all of OS, display device, displaying program, and user configuration
>choices.

As a step towards a Unicode-savvy Python, would it be an idea to define
Unicode versions of repr(), str(), ob.__repr__() and ob.__str__()?

eg.
  urepr()
  ustr()
  ob.__urepr__()
  ob.__ustr__()

'print' could then use these if it knows stdout is a Unicode-savvy output
stream.

Just





From bernhard at alpha1.csd.uwm.edu  Thu Nov 11 15:31:19 1999
From: bernhard at alpha1.csd.uwm.edu (Bernhard Reiter)
Date: 11 Nov 1999 20:31:19 GMT
Subject: ASCII delimited files
References: <942231502.1914033961@news.interbusiness.it> <382A240B.B247AD6B@python.net> <382A2801.FDE3CA09@python.net> <382A4330.833C4EBD@easystreet.com> <8EsW3.962$Xo.4938@news-server.bigpond.net.au>
Message-ID: <slrn82m9sn.ktp.bernhard@alpha1.csd.uwm.edu>

On Thu, 11 Nov 1999 05:39:16 GMT, Mark Hammond <mhammond at skippinet.com.au> wrote:
>Yes - a good CSV parser would make a very useful addition to the library!
http://eh.org/~laurie/comp/python/csv/
	has some attempts to build better csv support.
	by Laurence Tratt
(Of course not listed on the python.org site. :( )

	Bernhard

-- 
Research Assistant, Geog Dept UM-Milwaukee, USA.    (www.uwm.edu/~bernhard)
Free Software Projects and Consulting 			   (intevation.net)  
Association for a Free Informational Infrastructure              (ffii.org)



From gerrit.holl at pobox.com  Sat Nov  6 06:17:11 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Sat, 6 Nov 1999 12:17:11 +0100
Subject: smtplib.py how to ?
In-Reply-To: <38237da0.3456244@news.force9.net>
References: <38237da0.3456244@news.force9.net>
Message-ID: <19991106121711.A18313@optiplex.palga.uucp>

Colm Rafferty wrote:
> Dear all,
> I cannot insert text into  the headers To: From: CC: etc, of the email
> The script below ( the sample in the smtplib.py module) only sends all
> the text in the body of the email.


> How do I insert text in the headers?

What do you mean?

msg = '''From: A friend <gholl at my.org>
To: My brother <hholl at my.org>
Cc: My sister <eholl at my.org>
Subject: this is an example
X-Mailer: a python script
Date: Sat Nov  6 12:15:51 CET 1999
Reply-To: My father <dholl at my.org>

This is the body
'''

regards,
Gerrit.
-- 
linuxgames.nl.linux.org	 All about games on Linux in Dutch (under construction).
www.nl.linux.org	       The Dutch resource for Dutch Linux documentation.
www.nl.linux.org/discoverb		    Learn foreign words and definitions.
www.nl.linux.org/~gerrit/asperger	 I'm an Asperger, go here for more info.
www.nl.linux.org/~gerrit			    About me and pictures of me.



From ScherBi at BAM.com  Thu Nov  4 13:27:29 1999
From: ScherBi at BAM.com (ScherBi at BAM.com)
Date: Thu, 4 Nov 1999 13:27:29 -0500 
Subject: Extending Python dynamically
Message-ID: <3F8BA119B48AD11197470008C7A493CB98B643@HQOBGEX1>

> Thank you for your help. The method you suggested works perfectly!

You're quite welcome.

> However, I do not understand completely how the Python 
> interpreter finds out that it has to load an so file. Normally -- I
believed -- 
> when Python sees an "import xxx" statement it looks for a xxx.py, xxx.pyc
or 
> xxx.pyo file in the $PYTHONPATH directories or an xxx subdirectory having
an 
> __init__.py file, but not an xxx.so file. This means that if I add the
current 
> directory to the $PYTHONPATH (such as export PYTHONPATH='.' Python will
find 
> any arbitrary so file in the $PWD?

The true answers to these questions are at level lower than I am able to go
into in Python, primarily because I think the answers are in Python's c
source.
Anyway, I believe that Python, on execution of imort xxx, will load xxx.so
if, and only if, it has the proper interface, that is, if it has been built
as a Python extension.
Try it with any other .so file and see what happens.


> Anyway, I am very happy that I do not have to recompile 
> Python every time I add a new extension library (ie. an so file) :-).

Yes, while others may deride possible performance degradation of the shared
object interface, I have been known to ignore an extension or two because it
required me to rebuild Python, but I'm really out of my league here.

Bill



From Gareth.McCaughan at pobox.com  Thu Nov 25 15:41:40 1999
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: 25 Nov 1999 20:41:40 +0000
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <867lj85g4c.fsf@g.local>
	<383CC835.B585EBDD@Lugoj.Com>
Message-ID: <86r9he2tiz.fsf@g.local>

James Logajan wrote:

[I said:]
>>   - if/then/else works on *statements* (strictly, suites of
>>     statements) rather than on *expressions*, so I can't
>>     say
>>       print "You scored %s point%s" % (score, if score==1 then "" else "s")
>>     or anything like that. (Or, as some people prefer to put it,
>>     Python has no ? : operator.)
> 
> Thanks to Python's dynamic typing you can easily add the C "?:" expression
> to the language by defining a single function that will work with all types:
> 
> def ifexp(cond, trueVal, falseVal):
>    if cond:
>       return trueVal
>    else:
>       return falseVal
> 
> print "You scored %s point%s" % (score, ifexp(score == 1, "", "s"))

That has the wrong semantics in the presence of side effects.
(It's fine for my toy example, but then so is "score<>1 and 's' or ''".

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
sig under construction



From rgacote at AppropriateSolutions.com  Wed Nov 24 08:17:46 1999
From: rgacote at AppropriateSolutions.com (Raymond Cote)
Date: Wed, 24 Nov 1999 08:17:46 -0500
Subject: Python complaints
In-Reply-To: <005201bf367e$4901fef0$f29b12c2@secret.pythonware.com>
References: <XFMail.991124113221.mikael@isy.liu.se>
 <81gj3i$fe3$1@mach.vub.ac.be>
Message-ID: <v04003a01b46194dec84b@[192.168.24.43]>

Thomas Hamelryck <> wrote:
> I don't want to start the discussion about indentation again; it's no use
> since indenting blocks is here to stay. The point is that if you want to 
> write code that looks ugly or unreadable you will succeed in doing that,
> indentation or not. In many cases, using a variable indentation (depending
> on what construct you are using or the length of a line) leads to code
> that is visuably much more appealing.

I was sure this 'indenting thing' would be one of the two major issues that kept me grumbling about Python. However, after just a month or so of casual Python programming, I find myself becoming dissatisfied with extra work required by braces and begin/end in my day to day C and Pascal programming.

Was totally surprised how quickly I came to like the indenting.  

Just my 2cents. 

Ray

-------------------------------------------------------------------
Raymond Cote, President                 Appropriate Solutions, Inc.
www.AppropriateSolutions.com       rgacote at AppropriateSolutions.com
603.924.6079(v)  POB 458, Peterborough, NH 03458    603.924.8668(f)





From jlj at cfdrc.com  Wed Nov 17 14:26:47 1999
From: jlj at cfdrc.com (Lyle Johnson)
Date: Wed, 17 Nov 1999 19:26:47 GMT
Subject: Embedded Python 1.5.1 crashes under WIN32
References: <3832F6FB.68AAA96E@tu-bs.de>
Message-ID: <38330158.170604796@192.77.87.74>

See this entry in the Python FAQ list:

	http://www.python.org/doc/FAQ.html#8.7

On Wed, 17 Nov 1999 19:42:03 +0100, Sven Havemann
<s.havemann at tu-bs.de> wrote:

>Hi!
>===
>
>I've modified the example from Python-1.5.1/Demo/embed: 
>
>ftp://ftp.cg.cs.tu-bs.de/pub/cg/people/havemann/embeddedPython.tgz
>
>What I want to do is to call PyRun_SimpleFile from my main program. This
>works fine under Irix 6.5, but the same thing crashes inside
>PyRun_SimpleFile when compiled with VC6.0SP2 under Win32, which is a
>great pity. I have attached the demo.cpp file, and the complete
>(compiled) project is in the tarfile above. Does anybody know why it
>crashes??? 
>
>I wanted to build it as a plain Win32 console application, and I've
>renamed it to demo.cpp because I wanted it to be compiled as a C++ file. 
>
>I'd appreciate if you could Re: to me directly using the email below,
>because I can't follow the news regularly.
>
>TIA, greetings, Sven.
>
>--
>__________________________________________________________________
> dipl-inform. Sven Havemann        Institut fuer ComputerGraphik
> Odastrasse 6                      Rebenring 18
> 38122 Braunschweig - Germany      38106 Braunschweig - Germany
> Tel. 0531/2808955                 Tel. 0531/391-2108, Fax: -2103
> mailto:s.havemann at tu-bs.de        http://www.cg.cs.tu-bs.de
>
>
>=========================demo.cpp==================================
>#include <stdio.h>
>#include "Python.h"
>
>// #define SHOWPYFILE
>
>static char *argv0;
>
>void initxyzzy(); /* Forward */
>
>int main(int argc, char ** argv)
>{
>   argv0 = argv[0];
>   Py_Initialize();
>
>   initxyzzy();
>
>   PySys_SetArgv(argc, argv);
>
>   printf("Hello, brave new world\n\n");
>
>   PyRun_SimpleString("import sys\n");
>   PyRun_SimpleString("print sys.builtin_module_names\n");
>   PyRun_SimpleString("print sys.modules.keys()\n");
>   PyRun_SimpleString("print sys.argv\n");
>
>   char* filename="test.py";
>   FILE* myfile=fopen(filename,"r");
>#ifdef SHOWPYFILE
>   char line[200];
>   printf("Reading file %s\n",filename);
>   while(fgets(line,200,myfile)!=NULL) {
>      printf("Read: %s",line);
>   }
>#else
>   PyRun_SimpleFile(myfile,filename);
>#endif
>   fclose(myfile);
>
>   printf("\nGoodbye, cruel world\n");
>
>   Py_Exit(0);
>   return 0;
>}
>
>char *getprogramname()
>{
>        return argv0;
>}
>
>/* A static module */
>
>PyObject *xyzzy_foo(PyObject *self, PyObject *args)
>{
>        if (!PyArg_ParseTuple(args, ""))
>                return NULL;
>        return PyInt_FromLong(42L);
>}
>
>PyMethodDef xyzzy_methods[] = {
>        {"foo",         xyzzy_foo,      1},
>        {NULL,          NULL}           /* sentinel */
>};
>
>void initxyzzy()
>{
>        PyImport_AddModule("xyzzy");
>        Py_InitModule("xyzzy", xyzzy_methods);
>}
>==================test.py======================================
>print "this is from PyRun_SimpleFile"
>for i in range(100): print i*i,
>==================OUTPUT IRIX==================================
>Hello, brave new world
>
>('__builtin__', '__main__', 'array', 'fast_umath', 'imp', 'marshal',
>'math', 'operator', 'posix', 'signal', 'strop', 'struct', 'sys',
>'umath')
>['xyzzy', 'os', 'exceptions', '__main__', 'posix', 'sys', '__builtin__',
>'site', 'signal', 'UserDict', 'posixpath', 'stat']
>['./demo']
>this is from PyRun_SimpleFile
>0 1 4 9 16 25 36 49 64 81 100 121 144 169 196 225 256 289 324 361 400
>441 484 529 576 625 676 729 784 841 900 961 1024 1089 1156 1225 1296
>1369 1444 1521 1600 1681 1764 1849 1936 2025 2116 2209 2304 2401 2500
>2601 2704 2809 2916 3025 3136 3249 3364 3481 3600 3721 3844 3969 4096
>4225 4356 4489 4624 4761 4900 5041 5184 5329 5476 5625 5776 5929 6084
>6241 6400 6561 6724 6889 7056 7225 7396 7569 7744 7921 8100 8281 8464
>8649 8836 9025 9216 9409 9604 9801
>
>Goodbye, cruel world
>==================OUTPUT WIN32==================================
>-- Hello, brave new world
>
>('__builtin__', '__main__', '_socket', 'array', 'binascii', 'cPickle',
>'cStringI
>O', 'cmath', 'errno', 'imageop', 'imp', 'marshal', 'math', 'md5',
>'msvcrt', 'new
>', 'nt', 'operator', 'pcre', 'regex', 'rgbimg', 'rotor', 'select',
>'signal', 'so
>undex', 'strop', 'struct', 'sys', 'thread', 'time')
>['xyzzy', 'os', 'exceptions', '__main__', 'ntpath', 'strop', 'nt',
>'sys', '__bui
>ltin__', 'site', 'signal', 'UserDict', 'string', 'stat']
>['T:\\WORK\\HAVEMANN\\SW\\PYTHON-1.5.1\\DEMO\\EMBED\\VC60.exe']
>=====================** CRASH **================================
>================================================================
>__________________________________________________________________
> dipl-inform. Sven Havemann        Institut fuer ComputerGraphik
> Odastrasse 6                      Rebenring 18
> 38122 Braunschweig - Germany      38106 Braunschweig - Germany
> Tel. 0531/2808955                 Tel. 0531/391-2108, Fax: -2103
> mailto:s.havemann at tu-bs.de        http://www.cg.cs.tu-bs.de




From evan at tokenexchange.com  Mon Nov  8 09:43:24 1999
From: evan at tokenexchange.com (Evan Simpson)
Date: Mon, 8 Nov 1999 08:43:24 -0600
Subject: making the Python case to my dept.
References: <9911051740130D.02530@william.jinsa.org> <7GOU3.272$Ae3.194028@news-dal.corridex.com> <801p16$10s$1@vvs.superst.iae.nl>
Message-ID: <tlBV3.286$Ae3.202317@news-dal.corridex.com>

http://www.alice.org

It's scripted in Python, although at the simplest level you just manipulate
objects with the mouse, and at the second simplest you construct animation
scripts by choosing from pop-up menus.

Carel Fellinger <cfelling at iae.nl> wrote in message
news:801p16$10s$1 at vvs.superst.iae.nl...
> where can I find more on this Alice thing? my kids would love to play with
> 3-D animations! (I take it that scripting it is with Python)

Cheers,

Evan @ 4-am





From akuchlin at mems-exchange.org  Thu Nov 18 17:14:00 1999
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: 18 Nov 1999 17:14:00 -0500
Subject: Proposal: Python Info Collective (another ramble)
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <JlFX3.3659$s87.67930@dfiatx1-snr1.gtei.net> <slrn82unnr.mja.thor@localhost.localdomain> <80ubrg$tuj$1@hub.org>
Message-ID: <3dso23jvmf.fsf@amarok.cnri.reston.va.us>

T. Maddletyn at news.vex.net (T. Maddletyn) writes:
> It is true that Parnassus is horribly inefficient in some  areas (some
> i'm aware of, perhaps some i'm not aware of). However the good news is

It probably doesn't greatly matter, at this point; people will
complain if it gets too slow.  For now, the primary problem is getting
it to a host without a weekly bandwidth limitation.  Someone mentioned
putting it onto starship.python.net; is anything happening with that?

> Meanwhile the discussion continues. Formats. XSA, RDF, OSD, LSM, RFC-
> 822, on and on... i never really heard of any of these before! Well, 
> i'd heard of RDF, but didn't know really anything about it (I know 

Don't worry about formats; later on we can worry about writing scripts
that walk all over the database generating RDF or whatever.  First,
let's get it hosted somewhere suitable, then get more data into it,
and then worry about refinements such as improved performance or
alternative formats.

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
We all live to a formula. Maybe the secret lies in keeping that formula secret.
    -- Peter Greenaway, _Dear Boull?e_




From vroonhof at frege.math.ethz.ch  Sat Nov  6 16:44:40 1999
From: vroonhof at frege.math.ethz.ch (Jan Vroonhof)
Date: 06 Nov 1999 22:44:40 +0100
Subject: A newbie dictionary question
References: <byr9i42us9.fsf@bolzano.math.ethz.ch> <3824648F.7FB08EAC@x.x>
Message-ID: <bybt978fcn.fsf@bolzano.math.ethz.ch>

Eric Jacobs <x at x.x> writes:

> >     def __str__(self):
> >         self.value
> 
> Unrelated, but did you mean "return self.value"? It's
> legal to throw away values in Python.

Yes I did.

> >     def __cmp__(self,other):
> >         return self.nospaces() == other.nospaces()
> 
> The __cmp__ is returning the wrong thing. It should return
> -1 for <, 0 for ==, or 1 for >. You probably want
> 
>       def __cmp__(self, other):
>            return cmp(self.nospaaces(), other.nospaces())

Aaargghh. Of course. I feel stupid. Thanks a lot. I could have sworn
I wrote that. I guess it really was too late. It works fine now.

Jan



From herzog at online.de  Thu Nov  4 06:27:57 1999
From: herzog at online.de (Bernhard Herzog)
Date: 04 Nov 1999 12:27:57 +0100
Subject: __str__ vs. __repr__
References: <000401bf25b9$88ea57e0$2e2d153f@tim> <199911040157.UAA11908@eric.cnri.reston.va.us>
Message-ID: <m31za6339e.fsf@greebo.nodomain.de>

Guido van Rossum <guido at CNRI.Reston.VA.US> writes:

> It's a separate dilemma from the other examples.  My problem here is
> that I hate to make assumptions about the character set in use.  How
> do I know that '\237' is unprintable but '\241' is a printable
> character?  How do I know that the latter is an upside-down
> exclamation point?  Should I really assume stdout is capable of
> displaying Latin-1?  Strictly, the str() function doesn't even know
> that it's output is going to stdout.  I suppose I could use isprint(),
> and then Fran?ois could use the locale module in his $PYTHONSTARTUP
> file to make it do the right thing.  Is that good enough?  (I just
> tried this briefly.  It seems that somehow the locale module doesn't
> affect this?!?!
> 
> I still think that ['a', 'b'] should be displayed like that, and not
> like [a, b].  I'm not sure what that to do about a dict of Rat()
> items, except perhaps giving up on the fiction that __repr__() should
> return something expression-like for user-defined classes...
> 
> Any suggestions?

How about a hook function a la __import__? I.e. have a function
__format__ [1] in __builtin__ that is bound to repr by default. Whenever
Python prints the result of an expression in interactive mode, it calls
__format__ with the result as parameter and expects it to return a
string ready for printing.

If somebody want str instead they can put

import __builtin__
__builtin__.__format__ == str

in their $PYTHONSTARTUP file.

This scheme even allows for more fancy formatting like not escaping
printable chars in latin1 if the user wants it.


[1] the name __format__ is perhaps a bit too generic, but something more
descriptive like __interactive_repr__ may be a bit awkward, but then
again you don't normally call it manually.

-- 
Bernhard Herzog	  | Sketch, a drawing program for Unix
herzog at online.de  | http://www.online.de/home/sketch/



From T.Muddleton at news.vex.net  Mon Nov 15 13:13:05 1999
From: T.Muddleton at news.vex.net (T. Muddleton)
Date: 15 Nov 1999 18:13:05 GMT
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <14384.16479.836502.161209@anthem.cnri.reston.va.us>
Message-ID: <80pifh$12n1$1@hub.org>

On Mon, 15 Nov 1999 12:18:23 -0500 (EST), webmaster at python.org wrote:
>some editorial review for record changes, because the change isn't
>effective immediately.  That could be another bottleneck in the

Yup, i'm a bottleneck. Truer words were never spoken. Though the bottle
neck is not doing too badly today. Mailman update is processed.

I had considered (and still do consider) implementing a system
whereby owners can edit their own stuff (i see in the Trove shema
that they have planned for this already too)... but as i didn't know
how far my Parnassus scheming would get i didn't implement it yet.
Also since i have (until today) found and entered all the data 
myself I haven't put much thought into the other possible mechanisms
yet. There are other complication's i'll spare you. But I just want
to let you know that I have thought of this, and perhaps in the 
future the bottleneck can be removed, or at least widened a bit. 

>somewhere on the python.org site, but I'd still like Parnassus' author
>to get in touch with me...

I must say i'm a little worried about bandwidth issues after the
discussions here. But on the other hand... we'll never know what the
system can or can't take until we try right? Heh.

Or perhaps it would be prudent for me to look into possibilities
further.... i'm in a bit of a quandry. I've gotten more response than 
I expected (which has been nice... but a little frightening too!).

....




From hinsen at cnrs-orleans.fr  Thu Nov  4 05:28:47 1999
From: hinsen at cnrs-orleans.fr (Konrad Hinsen)
Date: 04 Nov 1999 11:28:47 +0100
Subject: Extending Python dynamically
References: <38215010.DCFCDBFA@iqsoft.hu>
Message-ID: <m3g0ym7dpc.fsf@chinon.cnrs-orleans.fr>

"N?meth Mikl?s" <nemeth at iqsoft.hu> writes:

> I have read ext.pdf (Extending Python with C or C++). As far as I
> understood (by reading chapter 2), I have to recompile my Python
> interpreter, and have to relink the interpreter against this DCOracle
> shared objects.
> Does Python not support a more dynamic extension mechanism (based on
> dlopen())? If I want to add an extension in C, shall I always

It does, at least for modern Unix variants and for Windows.
Check sections 2 (Unix) and 3 (Windows) of the current version of
the Extending&Embedding manual. The very first page of section 2
explains how to compile extension modules into shared libraries
under Unix.

> I have also started evaluating Zope. After the first two hours I found
> Zope is a fantastic product, I am curious about how people did dare
> implement a so complex application (ie. an application server) in an
> interpreted "scripting language". These type of software pieces are

Just don't call it a scripting language, maybe that helps!
-- 
-------------------------------------------------------------------------------
Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.55.69
Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais
-------------------------------------------------------------------------------



From smst at quantisci.co.uk  Thu Nov 18 06:03:46 1999
From: smst at quantisci.co.uk (Steve M S Tregidgo)
Date: Thu, 18 Nov 1999 11:03:46 GMT
Subject: __str__ vs. __repr__
Message-ID: <199911181103.LAA05706@cobweb.quantisci.co.uk>

Fredrik Lundh wrote:
> Phil Hunt <philh at vision25.demon.co.uk> wrote:
> > How about:
> > 
> >    x = a + b + c
> >    x = (a + b) + c
> >    x=a+b+c
> > 
> > I'd guess these all produce the same bytecode.
> 
> footnote: the abc80 computer I mentioned
> in my earlier post had an extra byte code
> for this purpose.  users can buy that you
> add spaces and reformat their code, but
> they won't accept that you remove their
> parentheses...
> 

Two solutions:

(1) When regenerating the source, wrap every single binary op with
parentheses; the bytecode generated by (a+b)+c will also be generated
by ((a+b)+c), and at least the user's original bracketing will be
there too.  Looks very messy though.

(2) Use some intelligence in deciding where brackets need to go. 
Clearly if (a+b)*c was compiled into bytecodes, we would have to
ensure the brackets remained, but (a*b)+c uses extra brackets that
can be ommitted during the decompiling stage.

This second solution wouldn't be so hard -- the order of the
bytecodes tells us which ops come first, and so we'd just need to
bracket a nested op if the op that uses the result is of a certain
type (so if a BINARY_ADD applied to elements a+b and c, there would
be no need to introduce protective parentheses, but there would for a
BINARY_MULTIPLY on a+b and c).

Anyhow, since decompiling should mainly be used to recover lost code,
as long as the code works the issue of (orignally extraneous)
brackets missing isn't really a problem -- I for one would just be
grateful that I got my code back.  (And heck, if I actually noticed
that the odd bracket was missing from a thousand-line file, I'd
consider that I'd got a bit _too_ familiar with the code ;-)

Cheers,
Steve



From dubois1 at llnl.gov  Sat Nov 13 11:42:59 1999
From: dubois1 at llnl.gov (Paul F. Dubois)
Date: Sat, 13 Nov 1999 16:42:59 GMT
Subject: Max Size of Array when using Numeric.py
References: <38286054.C8612EB5@edf.fr>
Message-ID: <nygX3.2040$V4.64389@news1.frmt1.sfba.home.com>

There is no intrinsic limit on the size of NumPy arrays. You just ran out of
memory, I guess.

Python 1.5.2 (#0, Apr 13 1999, 10:51:12) [MSC 32 bit (Intel)] on win32
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> from Numeric import *
>>> a=arange(5500*200)
>>> a.shape
(1100000,)
>>>

5500*200 = 1,100,000 * 4 = ~ 4MB per copy.
You have at least three copies, mv, eval(mv), and the resultant array.
eval(mv) is going to be double sized since the default Python real is 8
bytes. You aren't clear about mv but likely it is more than 4MB too. So
maybe we are pushing 16-20 MB. That doesn't really seem like enough for a
memory error.

You could write a C extension that created the array and did the reading of
each element and placing it into the array so that you only had the one
copy.

Helka FOLCH <helka.folch at edf.fr> wrote in message
news:38286054.C8612EB5 at edf.fr...
> I've tried to create an array object whose shape is (5500,200) but I get
> a memory error. Is there a maximum size for array objects?
>
> Maybe there's a better way of creating the array than what I did :
> My program reads the array which is in another format from a file. It
> transforms the array to the right format and holds it in a variable(mv)
> as a string. Then I just do:
>
> m = array(eval(mv),Float)
>
> Thanks a lot,
> Helka
>





From Gareth.McCaughan at pobox.com  Thu Nov 25 15:43:41 1999
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: 25 Nov 1999 20:43:41 +0000
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <867lj85g4c.fsf@g.local>
	<383CC835.B585EBDD@Lugoj.Com> <m3ogcjrn41.fsf@atrus.jesus.cam.ac.uk>
	<383D9C63.3A7E5806@Lugoj.Com>
Message-ID: <86ogci2tfm.fsf@g.local>

James Logajan wrote:

> Oh yeah? Well yer mudder wears army boots! Seriously though, the price I am
> talking about has absolutely nothing to do with simplicity. It has to do
> with the fact that languages such as Java that have "true" GC are
> non-deterministic in their execution times because GC can cause execution
> pauses in the VM at unpredictable times. This makes Java unacceptable for a
> large number of real-world applications that I have been involved in lately.
> ("I'm sorry, but your 911 call can not go through for another 20 seconds
> while our Java VM garbage collects 1GB of memory on the switch.") I have
> recommended Python over Java in at least one environment because of Python's
> deterministic approach to memory management. Had Python not had that, I
> suppose we would have gone with some other scripting language or even
> designed our own (gak!).

Garbage collectors have got a *lot* better since the days when
GC meant half-second pauses waiting for the machine to sweep
through all its memory. Let alone 20 seconds.

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
sig under construction



From phd at phd.russ.ru  Thu Nov 25 06:17:28 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Thu, 25 Nov 1999 11:17:28 +0000 (GMT)
Subject: Help with SMTP
In-Reply-To: <81hase$bdj$1@news.fsu.edu>
Message-ID: <Pine.LNX.4.21.9911251115450.11528-100000@fep132.fep.ru>

On Wed, 24 Nov 1999, Glenn Kidd wrote:
> everyhting looks in order but the website reports a "500 Internal Server
> Error".  When I look at the server log I get the following description of
> the error:
> 
> 
> [Wed Nov 24 13:23:30 1999] [error] [client 128.186.86.106] Premature end of
> script headers: /home/httpd/cgi-bin/mail.py

   Your script must return something to the server - either text or
redirect.

print "Content-Type: text/html"
print
print "<HTML>...and all HTML stuff...<BODY>The message was successfully sent</BODY></HTML>"

Oleg.
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From sekter at mail.matav.hu  Fri Nov 19 06:41:38 1999
From: sekter at mail.matav.hu (Arpad Kiss)
Date: Fri, 19 Nov 1999 12:41:38 +0100
Subject: Batch window
References: <383525CB.6E2EE3AF@obop.com.pl>
Message-ID: <006301bf3283$0c02d500$029f38c3@geometria.hu>

Change the .py extension to .pyw, then your Python program starts without
that window.
Arpad

----- Original Message -----
From: Przemys?aw G. Gawro?ski <gawron at obop.com.pl>
To: Python Mailing List <python-list at cwi.nl>
Sent: Friday, November 19, 1999 11:26 AM
Subject: Batch window


> Hello
>
> Is it posible to automaticaly minimize the python window, when starting
> a GUI application on Windows ?
>
> Thankx
>
> Przemek
> --
> Przemyslaw G. Gawronski UIN:8358522
> mailto:gawronskip at usa.net
> mailto:gawron at obop.com.pl
>
>
>
> --
> http://www.python.org/mailman/listinfo/python-list
>




From maytagrepairman at yahoo.com  Mon Nov 15 15:42:14 1999
From: maytagrepairman at yahoo.com (maytagrepairman at yahoo.com)
Date: Mon, 15 Nov 1999 20:42:14 GMT
Subject: Need help dealing with spaces in NT paths.
Message-ID: <80pr76$jq5$1@nnrp1.deja.com>

Sorry, I know this has to be a FAQ in another newsgroup somewhere since
this is really a Windows/DOS type question that I would think is really
common.

I'm trying to convert a script I wrote in Tcl to Python.  The Tcl
script is run only on UNIX (85% of my software development experience
is in UNIX BTW).  I'm developing the new script on an NT box and want
the script to work for both NT and UNIX.  I have a need to run various
compiled programs from the script.  In Tcl I simply used exec() to do
this.  The easiest way to do this in Python seems to be os.system().  I
don't know ahead of time where or what these external programs are.  In
fact they are passed in as arguments to the script.  I convert these to
absolute path names so I can run them from any directory I choose.
Unfortunately this doesn't seem to work when the executables have
spaces in the full path name.  This doesn't seem to be a Python problem
since I have the same problem when I type the commands in a DOS window.

Anybody have any advice for getting around this?


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



From b2blink at hotmail.com  Fri Nov 19 13:15:44 1999
From: b2blink at hotmail.com (Ulf Engstrøm)
Date: Fri, 19 Nov 1999 13:15:44 CET
Subject: Python Chinese
Message-ID: <19991119121544.5006.qmail@hotmail.com>

Is there any way to use Python with Chinese and other asian chars?
If not, are there any plans to make this compability?
Best regards
Ulf Engstr?m

http://www.b2b-link.com
ulf.engstrom at b2b-link.com

______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com



From greg.ewing at compaq.com  Wed Nov  3 03:37:39 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Wed, 03 Nov 1999 21:37:39 +1300
Subject: Builtin subclassing idea (Re: General q. `bout subclassing)
References: <381EE780.EBF18F59@bibsyst.no> <19991102151657.G29354@xs4all.nl>
Message-ID: <381FF453.84C6E5F9@compaq.com>

Thomas Wouters wrote:
> 
> I believe the classless-builtin-types problem is supposed to be fixed by
> Python 2.0. But I've seen that in response to so many ideas...

In this case, the problem is that nobody has thought of a way
of implementing it that wouldn't greatly slow down operations on
builtin types, or make builtin types take up a lot more memory,
or both.

Mr. Skaller in his Viper project seems to have hit upon a way
of solving part of the problem, allowing methods to be added to
the existing builtin types, and it seems that the same technique
could be used in CPython. It still doesn't let you make subclasses
of the builtin types, however.

Hmmm, I've just had an idea... what if the interpreter were
to recognise when you wrote something like

   class Furble(Listtype)
      ...

and behave as though you had instead written

   class Furble(UserList)
      ...

Then you'd be able to at least *pretend* that you can
subclass builtin types...

To implement this in a general way, add a "convert_to_class"
method slot to the TypeObject. Whenever an object which
isn't a class is used as a base class, its convert_to_class
operation would be invoked. For lists and dictionaries
this would import and return UserList or UserDict, and 
similar emulation classes could be added for the other 
builtin types.

Greg



From c.hintze at gmx.net  Tue Nov  2 15:17:52 1999
From: c.hintze at gmx.net (Clemens Hintze)
Date: Tue, 2 Nov 1999 20:17:52 GMT
Subject: Comparison between Python and "Ruby"
References: <1270663667-24391282@hypernet.com> <381ED7A6.D7D7FB13@compaq.com> <877lk1wiij.fsf@ev.netlab.co.jp> <19991102154543.A4469@optiplex.palga.uucp>
Message-ID: <m33duosl5b.fsf@qiao.localnet>

>>>>> "Gerrit" == Gerrit Holl <gerrit.holl at pobox.com> writes:

    Gerrit> Yukihiro Matsumoto wrote:
    >>  I guess he loves reducing variablity, if possible, which is
    >> totally against TMTOWTDI.  I don't judge whether it is good or
    >> bad.
    >> 

    Gerrit> Are there people who see TMTOWTDI [1] as an advantage?

Normally ... yes, I do! I do not like Python, because it only offers
one way. I don't try to avoid Perl, because it offers more than one
way. :-)

    Gerrit> [1] There's more than one way to do it.

[...]

Clemens



From nospam at bitbucket.com  Sat Nov 13 03:58:32 1999
From: nospam at bitbucket.com (Phil Mayes)
Date: Sat, 13 Nov 1999 00:58:32 -0800
Subject: method overriding trick
References: <14378.62498.649973.150075@bitdiddle.cnri.reston.va.us>
Message-ID: <lK9X3.696$An3.9333@newsfeed.avtel.net>

Jeremy Hylton wrote in message
<14378.62498.649973.150075 at bitdiddle.cnri.reston.va.us>...
>Overriding methods of a base class is a little clunky in Python.  You
>need to explicitly name the base class that implements the method, and
>you call its as an unbound method, passing self as the first argument.
>Example: Base.aMethod(self, arg)


and suggested using

>class Derived1(Base):
>    super_aMethod = Base.aMethod
>
>    def aMethod(self, arg):
>        return self.super_aMethod(arg) + 1
>
>...
>It doesn't solve the must-name-the-base-class problem, but it helps
>manage it better.  If you keep the override assignments at the top of
>the class statement, then you only have one place to look in each
>class when the inheritance hierarchy changes.


I use a method that doesn't remove the clunkiness, but localises the
hierarchy to one line:

_Parent = basedlg.BaseDlg
class AddressDlg(_Parent):
    def __init__(self, param):
        self.local = param
        _Parent.__init__(self, 2345)

By using the same name "_Parent" in all modules, less thinking is needed,
and cut'n'paste coding (what, me?) works better too.
--
Phil Mayes    pmayes AT olivebr DOT com
Olive Branch Software - home of Arranger
http://www.olivebr.com/






From donn at oz.net  Sat Nov 27 12:06:18 1999
From: donn at oz.net (Donn Cave)
Date: 27 Nov 1999 17:06:18 GMT
Subject: fcntl/ioctl and O_NONBLOCK
References: <slrn840205.sn4.hildeb@stahlw06.stahl.bau.tu-bs.de> <slrn83t5mv.bhk.hildeb@stahlw06.stahl.bau.tu-bs.de> <81ml5t$eer$0@216.39.151.169>
Message-ID: <81p32a$m14$1@216.39.151.169>

Quoth hildeb at www.stahl.bau.tu-bs.de (Ralf Hildebrandt):
| On 26 Nov 1999 18:57:01 GMT, Donn Cave <donn at oz.net> wrote:
|
| >I don't know about the ioctl, but you may not need it after all if
| >you're already going to use select.  Select will tell you if the
| >I/O would block, if so then don't read.
|
| Ah! So a read only block if there's nothing at all to read?

Right - if there's anything to read, you will read only that and
not block.  (So of course you must always check how much you actually
have read!)  Select tells you precisely that your I/O won't block,
either because there's data to read or end of file.

|> (And you can't read from both ends, "stdin" (i.e., Popen3.tochild)
|> is opened for write and attempts to read from it will probably be
|> frustrating.)
|
| :) Yup. I read the sockets HOWTO by now, and select() on a file/pipe 
| is more or less the same on unix. 

I guess - except of course, sockets are bi-directional, and pipes are
one direction only.

	Donn Cave, donn at oz.net



From da_cheng at my-deja.com  Thu Nov  4 14:47:57 1999
From: da_cheng at my-deja.com (da_cheng at my-deja.com)
Date: Thu, 04 Nov 1999 19:47:57 GMT
Subject: Installation Problem
Message-ID: <7vsntc$8qf$1@nnrp1.deja.com>

Hi. I downloaded the install files py152.exe and win32all-125.exe,
and installed them. Now Python command line works,
but IDLE (pythonw.exe) crashes on launch with an application error.

Also, the installation of win32all-125 had errors when
registering something, complaining sth about interface
not supported. with a strange looking number
-2147483646 in front of it.

Could someone please help?
thx
Da


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



From skip at mojam.com  Tue Nov  9 12:52:08 1999
From: skip at mojam.com (Skip Montanaro)
Date: Tue, 9 Nov 1999 11:52:08 -0600 (CST)
Subject: Production Web Deployment?
In-Reply-To: <NBBBKMMANMGEEBHPJJECGEGIFJAA.dirkl@home.com>
References: <199911080604.BAA22580@python.org>
	<NBBBKMMANMGEEBHPJJECGEGIFJAA.dirkl@home.com>
Message-ID: <14376.24392.828292.926057@dolphin.mojam.com>

>>>>> "Dirk" == Dirk Leas <dirkl at home.com> writes:

    Dirk> Is anybody taking lots of hits with httpdapy, PyApache,
    Dirk> ASP+Python, fastcgi, or Zope?

I use a combination of Apache and ZServer (from the Zope distribution) and a
separate backend database server on the Musi-Cal web server:

    Apache --HTTP proxy--> ZServer --XML/RPC--> db server

Apache handles several virtual servers, but the main one is Musi-Cal.  On
behalf of Musi-Cal it handled about 65K hits in the past two days.  Many of
those hits get proxied off to the ZServer (almost all html files are run
through DocumentTemplate and all search requests are passed off).  The
ZServer handled 21K requests in the past two days.  In addition, the
database server handles XML-RPC requests from a second machine on the same
physical network (www.mojam.com) that uses Mason (Apache + mod_perl + Mason
components).  Musi-Cal's database server handled about 160K requests of one
sort or another in the past two days (search URLs can result in several
XML-RPC calls and we do lots of other backend queries in support of other
activities).

I figure the Musi-Cal server is handling about 1.2M web hits per month right
now.  It's a Pentium II 400MHz w/ 128MB of memory running RedHat Linux 5.2
(264 bogomips according to /proc/cpuinfo).  The load average runs around 0.3
during the day with bursts above 1.0 every 15-30 minutes.  The hardware has
been up for almost 50 days.  I restart the software servers for various
reasons typically once or twice a day, however.

We have a more powerful dual processor machine (450+450 bogomips) that we
were using but set aside because of Linux SMP problems.  We've yet to buckle
down and address them.  Mostly we just wait for new stable revs of the
kernel and try again.  Given the performance of our current server, I don't
expect to be pressed to put it back into server for awhile yet.

Skip Montanaro | http://www.mojam.com/
skip at mojam.com | http://www.musi-cal.com/
847-971-7098   | Python: Programming the way Guido indented...




From co at kelly.daisybytes.su.uunet.de  Tue Nov 30 08:34:02 1999
From: co at kelly.daisybytes.su.uunet.de (Carsten Oberscheid)
Date: Tue, 30 Nov 1999 13:34:02 +0000
Subject: Finding Good Human Resource Candidates
References: <199911291502.YND4953@rciinfo.com>
Message-ID: <slrn847kia.e8o.co@kelly.daisybytes.su.uunet.de>

On Mon, 29 Nov 1999 3:02:59 PM -0500, News Reporter <newsreporter at rciinfo.com> 
spammed:
>The competition for finding good Human Resource candidates if fierce! 

Oh yef, it if... ifn't it?

> However, there is an edge - and we want you to have it FREE!

SREE? THREE? Parfe error....

oh, yeah, I *do* shut up, I mean, fhut up. Back to bored mode.


co.



From ivanlan at callware.com  Sat Nov  6 14:52:47 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Sat, 06 Nov 1999 12:52:47 -0700
Subject: making the Python case to my dept
Message-ID: <3824870F.7D34DA4C@callware.com>

Hi All--

Patrick Phalen wrote:
> 
> [Ivan Van Laningham, on Sat, 06 Nov 1999]:
> 

[snip]

> :: If there's a curve coming to Utah, I sure hope it will be one with jobs
> :: for Pythonistas, because there certainly aren't *any* here now.
> 
> Look at it this way: the HP jobs all appear to center around developing
> the E-speak technology. Python has been chosen as the official
> scripting language for E-speak. Novell has publicly endorsed E-speak. If
> Novell decides to builds Netware hooks into E-speak,  they'll probably
> hire rafts of Pythonistas in Utah.  :)
> 

Urk.  Trouble is, they're 50 miles from where I live!  Through miles and
miles of intense interstate reconstruction. ...

But. ... Hmm.  Hey out there, how is that NetWare port of Python coming?

<and-you-thought-it-was-hard-to-get-it-running-on-the-amiga>-ly y'rs,
Ivan;-)
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From coursesm at sbdhcp-4024.statenisland-ny.est.tcg.com  Mon Nov  1 13:53:05 1999
From: coursesm at sbdhcp-4024.statenisland-ny.est.tcg.com (coursesm at sbdhcp-4024.statenisland-ny.est.tcg.com)
Date: 1 Nov 1999 18:53:05 GMT
Subject: Feature request for interactive interpreter
References: <51C9F3C392D0D111A15600A0C99E64A9AF5606@localhost>
Message-ID: <slrn81rp2i.lkq.coursesm@sbdhcp-4024.statenisland-ny.est.tcg.com>

On Mon, 1 Nov 1999 12:03:35 -0700 , Bjorn Pettersen <bjorn at roguewave.com> wrote:
>I'm getting tired of typing, so my request for the next version of the
>interactive interpreter is a "help" function, probably defined as:
>
>	def help(x): print x.__doc__
>
>(with the appropriate 'no help available' message if the object doesn't have
>a doc string...)
>
Hmm, I do that already with a little script called startup.py.  In my
.bashrc is the following:
export PYTHONSTARTUP=~/startup.py

This causes startup.py to be executed every time the interpreter is started.
Hence, no need to muck with Python's source in any way.

I assume there is an analogue on non-*nix systems.

HTH,
Steve

>pretty-pleas'ly y'rs 
>--bjorn
>
>



From boud at rempt.xs4all.nl  Wed Nov  3 13:49:32 1999
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 3 Nov 1999 18:49:32 GMT
Subject: Need Berkley DB!!
References: <381D5E84.98931682@bibsyst.no> <hAQT3.488$bu3.39846291@news.randori.com> <381FE1D5.C0F721A6@bibsyst.no>
Message-ID: <7vq03s$ld$1@essle.valdyas.org>

Thomas Weholt <thomas at bibsyst.no> wrote:

> I think the problem might be solved if I, sigh, re-install Red Hat 6.1 and
> trash that oversized, over-featured Suse 6.2 Linux distro I got now. The Red
> Hat version of Python came with all that stuff installed. There`s not enough
> hours in a day to keep working on things like this. At least not right now.

Funny, that's exactly what I think of my current Suse setup... It was
dashed difficult just to get tkInter working with it. I'll be going
back to slackware, though, wanting something simple, yet KDE friendly.

-- 

Boudewijn Rempt  | http://www.xs4all.nl/~bsarempt



From mwh21 at cam.ac.uk  Wed Nov 24 14:25:20 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 24 Nov 1999 19:25:20 +0000
Subject: whitespace.py module
References: <y7b66yrdel7.fsf@agastya.eng.fore.com> <14396.14369.586465.200108@dolphin.mojam.com>
Message-ID: <m3r9hfsndr.fsf@atrus.jesus.cam.ac.uk>

Skip Montanaro <skip at mojam.com> writes:

>     Rajesh> I'm new to the python language, and the newsgroup. So, if this
>     Rajesh> newsgroup is not appropriate for this type of posting, please
>     Rajesh> pardon my ignorance and notify me.
> 
> Perfectly appropriate.
> 
>     Rajesh> I have written up a module called `whitespace.py' available at
>     Rajesh> <URL:http://www.dsmit.com/python/>
> 
>     Rajesh> and a sample program `whitespace' which uses this module, also available at
>     Rajesh> the same URL.
> 
> Have you checked out tabnanny in Tools/scripts/tabnanny.py of the Python
> distribution?  A merging of interests might be useful.  I suspect the two
> scripts detect similar problems.

I thought about posting that exact sentiment - but looked at the
website first. "whitespace.py" is designed to deal with text files; it
converts spaces to tabs, spots trailing spaces on lines, that sort of
thing. Not particularly useful to me, but not tabnanny either.

Cheers,
Michael




From wheinemanNOwhSPAM at uconect.net.invalid  Sat Nov 13 01:48:31 1999
From: wheinemanNOwhSPAM at uconect.net.invalid (wheineman)
Date: Fri, 12 Nov 1999 22:48:31 -0800
Subject: Win32 API Communication Functions Module
References: <09920fb9.076b8f52@usw-ex0102-013.remarq.com>
Message-ID: <1415c574.0fd5791a@usw-ex0102-013.remarq.com>

If you are having difficulties downloading, try entering the link by
hand. The HTML highlighting included the period. Dohhh.


* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!




From zippy at cs.brandeis.edu  Sun Nov 14 19:10:31 1999
From: zippy at cs.brandeis.edu (Patrick Tufts)
Date: Sun, 14 Nov 1999 16:10:31 -0800
Subject: Max Size of Array when using Numeric.py
References: <38286054.C8612EB5@edf.fr> <nygX3.2040$V4.64389@news1.frmt1.sfba.home.com>
Message-ID: <141119991610312514%zippy@cs.brandeis.edu>

I've created 6000 x 26000 arrays using Python 1.5.2 on solaris 2.6 x86,
and then done various matrix operations on them successfully.

In my limited working with float arrays, Pickle and cPickle have
problems dumping them to disk.  Otherwise, I've had no trouble.

--Pat



From conversy at lri.fr  Fri Nov 12 10:29:54 1999
From: conversy at lri.fr (Stephane Conversy)
Date: Fri, 12 Nov 1999 16:29:54 +0100
Subject: Tkinter and after
References: <382C2854.E9494EFE@lri.fr> <14380.11058.151680.518510@weyr.cnri.reston.va.us> <00a001bf2d1f$fcd3ac70$f29b12c2@secret.pythonware.com>
Message-ID: <382C3272.80EC434D@lri.fr>

excuse #1:
    someone has stolen my email adress, I did not write this one !
    nobody to believe me ?

excuse #2:

http://www.pythonware.com/library/tkinter/introduction/x8550-alarm-handlers-and-other.htm

    is not rendered by netscape the way it should, I missed the big 'after'
and took the little after that said tk blocks.
    you believe me now, do you ?

not a excuse #3:
    I apologize, I should read and test carefully.


Fredrik Lundh wrote:

> Fred L. Drake, Jr. <fdrake at acm.org> wrote:
> >  > It is said in Tkinter doc that after() is like a call to time.sleep()
> >  > wich blocks Tkinter.
> >
> >   I'm not sure what docs you're referring to (presumably the
> > eff-bot's), but after() does not block Tk.
>
> what version are you using?  if you
> run this script, can you move the
> scale widget after you've clicked
> the button?
>
> from Tkinter import *
>
> root = Tk()
>
> Scale(root).pack()
>
> Button(root, text="click me!",
>        command=lambda: root.after(100000)
>        ).pack()
>
> mainloop()
>
> </F>

--
St?phane Conversy
http://www-ihm.lri.fr/~conversy/
mailto:conversy at lri.fr






From mwh21 at cam.ac.uk  Tue Nov  2 10:33:12 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 02 Nov 1999 15:33:12 +0000
Subject: Printing strings
References: <381F00E1.721BB47F@cornell.edu>
Message-ID: <m3yacgc3if.fsf@atrus.jesus.cam.ac.uk>

Matthew Hirsch <meh9 at cornell.edu> writes:

> Hi All,
> 
> I'm trying to do something like the following:
> 
> a=4.0
> b=2.0
> print '2(%3.1f)' % a,b

ITYM

print 2*'%3.1f ' % (a,b)

the above is being parsed:

print (2(%3.1f)'%a),b
 
> where I really mean
> 
> print '%3.1f %3.1f' % a,b
> 
> but it's printing
> 
> 2(4.0) 2.0
> 
> Is there a way of doing this so I don't have to write out %3.1f,
> multiple times?
> 
> Thanks,
> Matt

hth,
Michael



From nikander at mindspring.com  Sat Nov  6 02:10:23 1999
From: nikander at mindspring.com (Rob Nikander)
Date: Sat, 06 Nov 1999 02:10:23 -0500
Subject: python apache module?
Message-ID: <3823D45F.CF78E6E7@mindspring.com>

I am just beginning to learn python and so far I think it is awesome.  I
was wondering if there is an Apache module analogous to mod_perl for
python?  I don't see anything on the Apache page.  If not, is someone
working on one?

Rob




From jeremy at cnri.reston.va.us  Tue Nov  9 14:45:35 1999
From: jeremy at cnri.reston.va.us (Jeremy Hylton)
Date: Tue, 9 Nov 1999 14:45:35 -0500 (EST)
Subject: __str__ vs. __repr__
In-Reply-To: <000101bf2a0b$2f87d0a0$e1a2143f@tim>
References: <u4glOIWXzGrqvJnI+IlPAQeurz1J@4ax.com>
	<000101bf2a0b$2f87d0a0$e1a2143f@tim>
Message-ID: <14376.31199.948463.211294@goon.cnri.reston.va.us>

>>>>> "TP" == Tim Peters <tim_one at email.msn.com> writes:

  [Toby Dickenson:]
  >> Tim's Rat class is a good example of how __repr__ should not be
  >> implemented, because it exposes the internal representation of
  >> the object, not the external value.

  TP> Beg to differ: it's an excellent example of how __repr__ should
  TP> be implemented.  

It's no accident that the function is called __repr__.  It's role is
to expose the representation.

Jeremy




From gvwilson at nevex.com  Thu Nov 25 12:17:18 1999
From: gvwilson at nevex.com (Greg Wilson)
Date: Thu, 25 Nov 1999 12:17:18 -0500
Subject: Cons (replacement for make) in Python?
Message-ID: <Bbe%3.8183$cz3.292454@nnrp1.uunet.ca>

Hi.  Has anyone built a tool like Cons (a pure-Perl replacement for make) in
Python?  If so, I'd welcome a pointer.
Thanks,
Greg






From fredrik at pythonware.com  Sat Nov 13 07:01:24 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 13 Nov 1999 13:01:24 +0100
Subject: __str__ vs. __repr__
References: <000001bf275c$02370d20$1d2d153f@tim> <7vvni7$19q0$1@nntp6.u.washington.edu> <382BEC79.EC65913D@maxtal.com.au> <382C6430.8DC07E3E@lemburg.com>
Message-ID: <01fd01bf2dce$d2c03cf0$f29b12c2@secret.pythonware.com>

M.-A. Lemburg <mal at lemburg.com> wrote:
> > You may be wondering how this works. The answer is that Viperi does not
> > use bytecode, it executes the 'Abstract Syntax Tree' directly.
> > Consequently, the original code can be recovered. (without comments,
> > or the original formatting, of course).
> 
> Could you point me to some resources ? Using ASTs for execution
> is an interesting subject and I would like to know how you deal
> with Python dynamic nature in this context (are the ASTs self
> modifying ?).

guess my brain doesn't work well today,
but maybe someone could tell me:

1) what's the reason AST's would have to change during
the execution of dynamic code (the byte code doesn't
exactly change, does it?)

2) what's the reason byte code cannot be used to recover
the original source code (Swedish readers may remember
the ABC80, which did exactly this).

</F>




From python-list at teleo.net  Sat Nov  6 16:12:35 1999
From: python-list at teleo.net (Patrick Phalen)
Date: Sat, 6 Nov 1999 13:12:35 -0800
Subject: A Modest Proposal -- <wink>
Message-ID: <99110613580407.02970@quadra.teleo.net>

As many of you know, on Wednesday November 3, Tim Berners-Lee sent the
XHTML 1.0 Proposed Recommendation back to the HTML Working Group for
further consideration regarding namespaces.

This unanticipated delay gives we Pythoneers a window of opportunity to
quickly submit a proposal to include a new XHTML element, namely The
Wink <wink>. In fact, with enough support, perhaps we could even get
<wink> codified as a fullblown XML element!

But to make this conformant, we will have to make one minor change.
Since <wink> has no matching closing element, it falls into the set of
EMPTY elements, of which the best known is <br>.

In XHTML, this has discrepancy been ameliorated by adding the closing
delimiter "/>" to "<br>", to yield "<br/>". This indicates to the
XML processor that the tag is EMPTY. Therefore, under the new proposal,
"<wink>" would, of necessity, become "<wink/>".

I'm sure everyone will agree that this small change will be more than
worthwhile, in order to make <wink> (not to mention
comp.lang.python itself) fully buzzword compliant.

Please consider writing email to the HTML Working Group in support of
this proposal.





From ANTIGEN_GIBRALTAR at captura.com  Thu Nov 25 17:40:52 1999
From: ANTIGEN_GIBRALTAR at captura.com (ANTIGEN_GIBRALTAR)
Date: Thu, 25 Nov 1999 14:40:52 -0800
Subject: Antigen found W32/Ska.A.Worm virus
Message-ID: <7309F1998D14D3119DD80008C7C9CA806C6796@GIBRALTAR>

Antigen for Exchange found Happy99.exe infected with W32/Ska.A.Worm virus.
The file is currently Deleted.  The message, "Newbie Question:  Generating
Timer Events wit Tkinter", was
sent from Jim Bishop and was discovered in Gabe Newcomb\Inbox
located at Captura Software, Inc./CAPTURA/GIBRALTAR.



From tbryan at python.net  Sat Nov  6 05:00:35 1999
From: tbryan at python.net (Thomas A. Bryan)
Date: Sat, 06 Nov 1999 05:00:35 -0500
Subject: making the Python case to my dept.
References: <9911051740130D.02530@william.jinsa.org>
Message-ID: <3823FC43.CF7FBD71@python.net>

William Ahern wrote:

> Could anybody suggest some substantive info to make the case (or not make it).
> Also, the department will need an IDE (Win95/Mac), so if anybody knows where to
> find one....

I discussed this with a CS student once.  He thought that it was 
better to start with a procedural (subset of a) strongly typed 
language: Pascal, C, or a subset of Java or C++.  He thought it a 
bad idea to start with such a loosely typed language as Python.  
His argument was essentially that students who first learned 
Python would find it very difficult to understand why they had 
to declare variables as a specific type, etc.  Then again, that's
how both he and I started out programming (with Pascal).

I started to convince him otherwise.  I think that it may be better 
to introduce students to programming by letting them think about 
problem decomposition and data type design.  With Python, students 
can be quickly introduced to basic concepts such as loops and 
procedures.  They can even probably work with user-defined objects 
in their very first course.  They could implement a tree structure 
as nested Python lists, for example.  They could be introduced to 
objects in their first course if desired.  For a student who has never 
programmed before, having to mess with static types often seems 
strange.  It certainly slows down a course and gives students an 
additional thing to deal with while their trying to learn to think 
like a computer, decompose problems, design consistent interfaces, 
encapsulate data, etc.  Many of us forget how much there is to learn 
in a beginning programming course after we've been programming for 
a while.

Python also has a large library (like Java) so that students can be 
assigned problems that *do* something very early in their courses.
Assignments aren't restricted to things like a four-function calculator
because students can lean on the library and start implementing more 
realistic applications as/after they learn the fundamentals of 
programming.

The interactive interpreter also enables students to test little 
language features while they're programming.  They can keep a window 
with the interpreter running open while they enter the program source 
code in another file.  Then, if they can't remember the methods for 
a list, they can do something like this

>>> L = []
>>> dir(L)
['append', 'count', 'extend', 'index', 'insert', 'pop', 'remove',
'reverse', 'sort']
>>> L.append(1)
>>> L
[1]
>>> L.append([1,2,3])
>>> L
[1, [1, 2, 3]]
>>> L.extend(4)
Traceback (innermost last):
  File "<stdin>", line 1, in ?
TypeError: list.extend() argument must be a list
>>> L.extend([4])
>>> L
[1, [1, 2, 3], 4]
>>> L.extend([5,6,7])
>>> L
[1, [1, 2, 3], 4, 5, 6, 7]
>>> print L.append.__doc__
L.append(object) -- append object to end
>>> print L.extend.__doc__
L.extend(list) -- extend list by appending list elements


Unlike Perl or Tcl, programming in Python doesn't look so much different 
from programming in C++ or Java.  It wouldn't be too difficult to 
transition from Python to either C++ or Java.  It would be simple to 
introduce the second language, too.  For example, begin by motivating a 
C++ course by explaining that unlike Python, their first language, C++ 
is compiled to a machine dependent executable.  Explain that the point
is
to make a very fast executable.  To permit the compiler to make 
optimizations, programmers are required to help it out by telling it
what 
"type" all the variables are.  This is similar to Python types, except 
that C++ variables can only (basically) hold one type during their 
lifetime.  The compiler must also resolve references at compile time.
Thus, the language gains speed by sacrificing some of Python's dynamic
features.  Then again, the C++ compiler provides type safety and catches 
a lot of bugs at compile time instead of at run time.  C++ is also
designed 
for very large programs where you might want to guarantee that others 
don't touch your object's implementation.  C++ provides very strong 
language features to separate an objects' implementations from their 
interfaces.  Explain why this separation is often a good thing.

The first half a day could then be a whirlwind intro to what C++ 
requires and provides.  The point is that after a semester or two of 
Python, students are hopefully competent programmers.  They know 
how to handle loops and write procedures.  They've also worked with 
objects and thought about the benefits of consistent interfaces and 
seen the benefits of subclassing to specialize behavior.  Thus, a 
whirlwind introduction of C++ could show them objects and how 
subclassing in C++ looks.  The potentially difficult concepts of 
object-oriented design were taught without the additional obstacles 
presented by a language like C++ or Java.  When learning C++, the 
students  would already understand the road map.  They understand 
objects; they would just be learning how objects fit in a statically 
typed language.  When they get annoyed with the rigidity of types, 
they'll be happy to learn templates to gain some of the lost 
flexibility.

> I know Python is being used on a broad range of applications. 
> Complaints from students these days, a professor told me, is 
> that the current language, Pascal, is outdated and doesn't 
> have much applicability in the real world. 

I thought Pascal was designed as a teaching language.  It was meant 
to make an introduction to programming more friendly than throwing 
the students to something like C.  Of course, some companies started 
implementing in Pascal, but that wasn't the point.  The point was 
to start students with a language designed to make learning to 
program easier.

If they originally chose Pascal for this fact, then they may 
like Python.  Guido designed a language that would be simple 
enough to teach everyone but powerful enough to implement real 
applications.  Go to the Python website to find a few links 
to where Python's being used in the "real world."

Still, few people are knocking down my door for Python programmers 
while there are plenty of jobs for C++/Java/Delphi programmers.
Then again, if there are plenty of jobs for Cobol programmers, 
and I doubt that anyone would advocate starting the students with 
that language.  The point here is that professors shouldn't worry 
about teaching students a critical job skill in their first CS class.
Do math majors or engineers learn their critical job skill's in their 
first classes?  No.  They don't often go into the job market saying, 
"I know calculus."  The skills they learn in their senior classes 
are the ones that get them the jobs (that and internships).  The first 
courses are there to lay a solid foundation.  If the students don't 
understand that, then they need to be told.  Again, there are jobs 
in distributed databases and CORBA, but I don't think either of these 
topics are suitable for an intro course.

Python would probably make a good first language because that was 
one of its design goals.  Point the professors to the Python website 
where they can read papers by Guido van Rossum, the language's 
designer.  Heck, Guido would probably be willing to correspond with 
them about how the language would fit in their curriculum.

> I've seen all of the energy and spirit behind the Python community, 
> and I can't fathom why any cash strapped organization would drop a 
> few grand for something that is freely available on the internet. 

Java is also freely available.  Other companies would probably be 
willing to waive some or all of the fees for student programming tools
if it guaranteed a whole graduating class would know their tools.

> I figure using Python would be a start to getting people to think 
> outside of the Microsoft culture (for better or worse).
> Bill Ahern
> MIS, JINSA

Ah!  Well, I assumed that you meant a CS department.  I see that 
you probably meant the MIS department...which also explains a 
market-based Microsoft focus.  I'm sure that any decent business 
student must understand the risk of trusting a single supplier, 
but I can see where the complaints about learning Pascal could 
come from.  I'm not sure how those same people would feel about 
learning Python since there still isn't much hype about it.

A bonus for Python though is that it can be used to interface 
Microsoft's COM components (thanks to Mark Hammond).  For 
scripting COM components, I'm sure professors would see the 
advantage of using a well-designed language like Python instead 
of the mess I hear Visual Basic is.  If that doesn't satisfy 
market-driven students, I don't know what will.  There is also 
a cross-platform Python IDE called Idle and a separate Windows 
IDE called PythonWin.

Another argurment for Python is that there are many tasks for 
which something like Java is overkill.  That's where languages 
like Python, Perl, Tcl, and Visual Basic fit.  I think that it's 
critical for students to know something about these languages 
since I haven't been in any job where at least one of them is 
not used.  Of the languages listed above (of which I can only 
claim to know Python and Perl), I'd say that only Python would 
make a good introductory language.  By starting with Python, 
the department would simultaneously lay the foundations for 
other programming courses and introduce students to the type 
of language that is used as a "glue" language...something a 
well-designed curriculum may address more fully in a later 
course.

Good luck.  I'd love to hear that the department chose Python.
---Tom



From smst at bigfoot.com  Mon Nov 15 11:56:27 1999
From: smst at bigfoot.com (Steve Tregidgo)
Date: Mon, 15 Nov 1999 16:56:27 GMT
Subject: Memory not being released?
Message-ID: <80pdvo$99r$1@nnrp1.deja.com>

 Hi there,

I'm having problems with a large process, and I'd really appreciate
some help in tracking down the cause of its bloatedness.

I have a process (goes up to 250+ MB, cue lots of virtual memory
usage and hence slowdown), which busies the server so much that it just
can't bring itself to do other things -- in other words it's severely
impractical.

The actual data that results from the process is relatively small (the
useful stuff is less than 1MB, and other data -- such as objects that
have been dealt with, to prevent following circular references -- is no
more than a few MBs), and certainly well below the size of the actual
footprint.

I've gone through the loops and recursion with a fine-tooth comb, and
found that by eliminating certain function calls the footprint is
reduced -- the process is thus rendered useless (they were the
important functions, of course), but at least I get an idea of where
the memory is going.

On the other hand, commenting out some other expressions (such as those
that remember data -- in other words, the things that I would expect to
take up memory and not release it again) doesn't make a blind bit of
difference.

So what's going on?  My guesses so far have been along the lines of
refcounts not going down -- how else do I explain why memory that was
allocated was seemingly not deallocated?  I've tried del'ing everything
in sight, storing ids instead of objects, inserting the odd
sys.exc_traceback=None to clear those pesky traceback reference holders
... none of it has helped so far.

Are there other hidden things that might hold references to my
objects?  And if so, what can I do about them?  After using an object
once, I don't want (don't need) to keep it in memory any more, so for
those objects I'll be revisiting at a later date, I just remember their
id and delete them.

If it's any help, the process does something like this:

def recurse(obj, list, dict1, dict2):
  # Prevent circular recursion...
  if dict1.has_key(obj.__id__):
    return
  else:
    dict1[obj.__id__] = None

  a = do_thing(obj, dict2)

  # Recurse into obj's children, or do something else
  for item in obj.subs:
    if item.spam:
      recurse(item, list, dict1, dict2)
    else:
      b = do_thing(item, dict2)

  c = do_thing(obj, dict2)

  # Remember some things -- commenting
  # out produces no memory saving.
  list.append(a,b,c)

The function do_thing implements a loop; in total (over the whole
process) the loop's body is executed tens of thousands of times -- it
causes the biggest memory saving when commented out (and of course it's
the most important bit), but does no obvious "remembering" of things.

Does anybody have any ideas?  Or can you maybe point me to some
documentation that deals with this -- the FAQ looked promising but
ultimately didn't help.

Cheers,
Steve

--
       -- Steve Tregidgo --
Developer for Business Collaborator
        www.enviros.com/bc


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



From htrd90 at zepler.org  Sun Nov  7 09:13:20 1999
From: htrd90 at zepler.org (Toby Dickenson)
Date: Sun, 07 Nov 1999 14:13:20 +0000
Subject: __str__ vs. __repr__
References: <000401bf25b9$88ea57e0$2e2d153f@tim> <199911040157.UAA11908@eric.cnri.reston.va.us> <m31za6339e.fsf@greebo.nodomain.de> <5logd98v8z.fsf@eric.cnri.reston.va.us>
Message-ID: <GYAlODECGAP0PXj92u7+Fxk0uRUP@4ax.com>

Guido van Rossum <guido at cnri.reston.va.us> wrote:

>Bernhard Herzog <herzog at online.de> writes:
>
>> How about a hook function a la __import__? I.e. have a function
>> __format__ [1] in __builtin__ that is bound to repr by default. Whenever
>> Python prints the result of an expression in interactive mode, it calls
>> __format__ with the result as parameter and expects it to return a
>> string ready for printing.
>
>Brilliant!  The default function could do all the current built-in
>magic -- print nothing if it's None, assign it to builtin _, and so
>on.
>

...and an IDE could open an object browser window. Mmmmmmmmm.



Toby Dickenson



From culliton at clark.net  Thu Nov 11 17:00:41 1999
From: culliton at clark.net (Tom Culliton)
Date: Thu, 11 Nov 1999 22:00:41 GMT
Subject: Beginning programming with Python
References: <sng0ychgwt.fsf@motorola.com>
Message-ID: <d0HW3.36459$oa2.141777@iad-read.news.verio.net>

In article <sng0ychgwt.fsf at motorola.com>,
Aaron Ginn  <aaron.ginn at motorola.com> wrote:
>I'm starting a non-profit organization dedicated to promoting computer
>literacy in low-income youth (ages 12-18).  The curriculum will
>include an introduction to programming using either Python or Perl.

A worthy undertaking.

>First, I know Perl very well, but I know nothing about Python.  I've
>heard from many sources that Python is the best language to teach
>someone with no programming experience because it emphasizes
>fundamentals and the correct (i.e. best) way to program.  Could anyone
>comment on how the two languages compare in this aspect?  Also, I will
>need to learn Python as well if I choose to go that route.  I am very
>literate with Perl and Tcl/TK.  How quickly can I expect to pick up
>Python?

A day for the basics and to be able to do useful stuff, a week to be
pretty comfortable.  I've taught quite a few people now, starting with
myself ;-), and it goes very fast.  Having someone who can look over
your shoulder, answer questions, and suggest better idioms speeds
things up, but if you've got a few procedural langauges under your
belt already, it's a slam dunk.

Two concepts to pay attention to which ease the process are the
mutable vs. immutable distinction and "everything is a reference
(object)".

>Secondly, if I choose to use Python for this, I would like to supply
>each student with a reference book on Python.  I've narrowed it down
>to a choice between the two O'Reilly books, "Programming Python" and
>"Learning Python".  PP apears to have a lot more meat to it, but LP
>appears to be more suited to the beginner.  Assuming the audience
>includes kids that have never programmed before, should I go with LP,
>or will I miss out on anything by not getting the more detailed book?

"Programming Python" is good book but it's more of a catalog/reference
than a learning book.  "Learning Python" is a very nice tutorial.
There are also several free tutorials available for folks with varying
levels of experience.  I learned from the standard tutorial that comes
with the on-line Python Documentation and thought it was excellent for
someone with a bit of experience.



From sragsdale at my-deja.com  Tue Nov 30 14:05:41 1999
From: sragsdale at my-deja.com (sragsdale at my-deja.com)
Date: Tue, 30 Nov 1999 19:05:41 GMT
Subject: Removing null items
Message-ID: <821761$s8u$1@nnrp1.deja.com>

So I've got a very simple desire: to remove all false (I.E. None, 0, '',
[]) items from a list.

As it is I've got this function which works fine, but it offends my
sight.  Is there any better (or at least different) way to do this?

#################################################################
# return non-null items from the first list
def foo(lst):
    rlist = []
    for i in lst:
        if i:
            rlist.append(i)
    return rlist


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



From aurag at geocities.com  Wed Nov 24 20:48:14 1999
From: aurag at geocities.com (Hassan Aurag)
Date: Thu, 25 Nov 1999 01:48:14 GMT
Subject: Newbie: raw_input and other input methods
Message-ID: <19991125.1481400@adam-aurag.penguinpowered.com>

 Hello,

This is my first question to this newsgroup.

I am writing a small math app using pygnome, python.

It is a terminal based app. 

I read input process and print output. It is based on a hybrid onecmd 
and InteractiveInterpreter from code.py.

I am using python 1.5.2.

Anyway, I used raw_input to be able to use readline and its nifty back 
arrow, up arrow...

However when I wanted to have a nifty prompt like the following one:

raw_input("\033[0;34mInput: \033[0m"), I was surprised it got all 
screwed up.

This is true if you run this anywhere (gnome-terminal, vt without 
X...). It won't know the real size of terminal, because it thinks 
len("\033[0;34mInput: \033[0m") = 24 instead of what is really there 
and that is 7.

This means that line wrapping won't work on your first line!

My question is:

Is there another input method a la raw_input (not like input that 
evals also) that can be mixed with readline and still sees that string 
as it is.

Or, is there any workaround?

Btw, the weird escape sequences turn prompt to blue!


Thanks in advance


H. Aurag
remove.aurag at geocities.com









From nbecker at fred.net  Wed Nov 10 09:19:43 1999
From: nbecker at fred.net (nbecker at fred.net)
Date: 10 Nov 1999 09:19:43 -0500
Subject: [graphing] Simple xmgrace python interface
Message-ID: <x88d7tifmyo.fsf@rpppc2.hns.com>

Actually, interfacing pyton to grace is really very simple.
You can get the newest information about Grace and download
the latest version at the 
http://plasma-gate.weizmann.ac.il/Grace/

-------------- next part --------------
A non-text attachment was scrubbed...
Name: grace.py
Type: application/octet-stream
Size: 534 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/19991110/f37fa28e/attachment.obj>

From mhammond at skippinet.com.au  Wed Nov 17 17:32:55 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 17 Nov 1999 22:32:55 GMT
Subject: Python in Windows CE?
References: <3832D94D.50A25475@nrel.colostate.edu>
Message-ID: <r2GY3.18$VA6.189@news-server.bigpond.net.au>

Rick Shory wrote in message <3832D94D.50A25475 at nrel.colostate.edu>...
>  Has anybody used Python for development under Windows CE?

Check out my CE pages at http://starship.python.net/crew/mhammond/ce/  Also
note the reference to the mailing list - you have a much better chance of
contacting the people who know about this stuff via that list.  Im not doing
much CE work lately, so that list is better than mailing me directly...

Mark.





From greg.ewing at compaq.com  Wed Nov 24 05:54:10 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Wed, 24 Nov 1999 23:54:10 +1300
Subject: Python Type-Inference based LINT.. (pylint.py)
References: <19991121180043.B3184@teapot.egroups.net> <81e2r7$j14$1@newshost.accu.uu.nl>
Message-ID: <383BC3D2.98BCD1FF@compaq.com>

Martijn Faassen wrote:
> 
> Note that this doesn't need to be a syntax extension; check out some of
> my Swallow posting with dejanews for examples of how it could work in
> current Python.

I think I'd prefer it if it *were* a syntax extension.
The suggestions I've seen so far for non-syntax-extending
type declarations strike me as rather ugly.

That's a minor point, though -- the main thing is to
get some sort of type checker to work! The syntax can
be thrashed out later.

> I myself would take the Python subset approach
> (i.e. Swallow) to make sure no runtime dynamics exist that can mess up
> any type checking. Seems to me the easier approach

Me, too. Also, I tend to think that explicit type
declarations are a valuable way for the programmer to
state the *intent* of the code in a way that both
humans and compilers can process. In my experience,
type errors are found sooner and reported in a more
easily understood way when explicit type declarations
are used.

I've tried it both ways when using Haskell-like
languages, and I find that the error messages I get
when I don't explicitly declare types tend to be of
the form

   Couldn't unify <some-big-hairy-complicated-type-
   expression> with <some-other-big-hairy-complicated-
   type-expression-that-looks-vaguely-similar-although-
   there-must-be-some-subtle-difference-if-only-I-can-
   find-it>.

After getting a few of those I quickly learned to
provide type declarations for ALL my functions...

Greg



From da at ski.org  Tue Nov  9 14:18:08 1999
From: da at ski.org (David Ascher)
Date: Tue, 9 Nov 1999 11:18:08 -0800 (Pacific Standard Time)
Subject: Fastest GUI toolkit?
In-Reply-To: <01bf29f1$cbfd5e60$5a41c3d1@Cyberus.cyberus.ca>
Message-ID: <Pine.WNT.4.04.9911091114510.251-100000@rigoletto.ski.org>

On Mon, 8 Nov 1999, Gordon Williams wrote:

> I have an application where I need to draw many lines fast and repeatedly
> (an oscilloscope).  The number of lines ranges from about 1000 to 16000. 
> Currently I have the application using Tkinter but I am finding it too
> slow.  I am currently using W95 as an os, but would prefer to make the code
> portable across other os if possible.  Right now speed is more important
> than portability so I would be willing to give this up.
> 
> I would like to draw 1000 lines in 1/10 of a second on a P166 machine.  (Am
> I way off the mark??)  I was hoping that someone would have tried different
> packages and could give me pointers in the right direction.

I'd use OpenGL w/ NumPy arrays.  Especially if you don't need antialiasing
(I'm guessing that these are horizontal or vertical lines), you could do
several hundred lines per frame (1/60th or so), possibly many more if you
have hardware acceleration.

MGL is another possibility, but it's far less portable, and the Python
bindings aren't nearly as stable (or released =).

--david

See: http://starship.python.net:9673/crew/da/Code/PyOpenGL
    




From gvwilson at nevex.com  Tue Nov  2 14:03:41 1999
From: gvwilson at nevex.com (Greg Wilson)
Date: Tue, 2 Nov 1999 14:03:41 -0500
Subject: Python + JavaSpaces?
Message-ID: <EAGT3.3546$cz3.100778@nnrp1.uunet.ca>

Has anyone built (or thought about building) a Python interfaces to
JavaSpaces?
Thanks,
Greg






From gvwilson at nevex.com  Thu Nov 25 12:31:17 1999
From: gvwilson at nevex.com (Greg Wilson)
Date: Thu, 25 Nov 1999 12:31:17 -0500
Subject: DDD and Numerical Python
Message-ID: <Coe%3.8191$cz3.292560@nnrp1.uunet.ca>

Hi.  Has anyone integrated DDD's data visualization with the Numerical
Python extensions?
Thanks,
Greg






From matz at netlab.co.jp  Tue Nov  2 12:49:54 1999
From: matz at netlab.co.jp (Yukihiro Matsumoto)
Date: 03 Nov 1999 02:49:54 +0900
Subject: Comparison between Python and "Ruby"
References: <1270663667-24391282@hypernet.com>
 <381ED7A6.D7D7FB13@compaq.com> <877lk1wiij.fsf@ev.netlab.co.jp>
 <003901bf251c$290775d0$f29b12c2@secret.pythonware.com>
Message-ID: <87zowwvl4t.fsf@ev.netlab.co.jp>

"Fredrik Lundh" <fredrik at pythonware.com> writes:

|> I guess he loves reducing variablity, if possible, which is totally
|> against TMTOWTDI.  I don't judge whether it is good or bad.
|
|you're confusing syntax with semantics.

I don't think so.  I don't believe anyone can reduce semantic
variability.  Thus TMTOWTDI applies to semantics AND syntax.

But it's OK.  I'm not Larry Wall nor Perl freak.

I just felt weird when Pythoneers say about TMTOWTDI or even
TMTOWTHMTOWTDI. 
						matz.



From gmcm at hypernet.com  Tue Nov 30 14:33:36 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 30 Nov 1999 14:33:36 -0500
Subject: WinNT and python as text filter...
In-Reply-To: <816010E2456BD111A48700805FBBE2EEFDF877@ex-quebec-u1.baan.com>
Message-ID: <1268132071-35992705@hypernet.com>

Gaetan Corneau wrote:

[Mark Hammond, to Tim Ottinger]
> > >If so, there is a bug in the NT command prompt - you need to 
> > specify the
> > >full path to Python.exe on the command line, and it will
> > >work...
> 
> This is exactly the problem I have reported a few months ago. I
> have access to many tests machines, and two of them had this
> problem from time to time. After a reboot, it disappeared, and I
> haven't seen it again since :)

Are you sure it's the same problem?

As far as I know
 >xxx.yyy >foo.txt
does not work if xxx.yyy is executed through a file 
association. foo.txt will be created and empty. The executable 
associated to .yyy will have stdout and stderr connected to a 
bit bucket. Python doesn't have anything to do with it.

- Gordon



From frehe491 at student.liu.se  Sat Nov  6 01:43:28 1999
From: frehe491 at student.liu.se (Fredrik Henbjork)
Date: Sat, 6 Nov 1999 07:43:28 +0100
Subject: python apache module?
References: <3823D45F.CF78E6E7@mindspring.com>
Message-ID: <99110607444800.15295@o112>

On Sat, 06 Nov 1999, you wrote:
> I am just beginning to learn python and so far I think it is awesome.  I
> was wondering if there is an Apache module analogous to mod_perl for
> python?  I don't see anything on the Apache page.  If not, is someone
> working on one?
> 
> Rob

http://www.msg.com.mx/pyapache/

--
Fredrik Henbjork

No victim, no crime. Support freedom of mind! 
http://www.maps.org | http://www.norml.org



From fdrake at acm.org  Wed Nov 17 10:22:46 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Wed, 17 Nov 1999 10:22:46 -0500 (EST)
Subject: Final classes like in java???
In-Reply-To: <14385.49269.629053.250222@goon.cnri.reston.va.us>
References: <80sao7$1848$1@rtpnews.raleigh.ibm.com>
	<14385.45862.55010.146256@weyr.cnri.reston.va.us>
	<14385.48236.202756.297241@goon.cnri.reston.va.us>
	<14385.48747.331513.47679@weyr.cnri.reston.va.us>
	<14385.49269.629053.250222@goon.cnri.reston.va.us>
Message-ID: <14386.51270.415682.535928@weyr.cnri.reston.va.us>

I wrote:
 >   That's exactly why I said it wasn't guaranteed.  ;-)

Jeremy Hylton quipped:
 > Would you be interested in buying my general purposes NP algorithm
 > solver?  It provides polynomial time solutions to all problems in NP.
 > I don't guarantee the results, but it should be good enough <wink>.

  Sure!  I'll pay you all the money I got for my suggestion on "final"
classes.  ;-)


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From greg.ewing at compaq.com  Mon Nov  8 07:33:46 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Tue, 09 Nov 1999 01:33:46 +1300
Subject: What are Aspects? (Re: How can I make my assertions smarter?)
References: <Pine.WNT.4.05.9911032042440.109-100000@david.ski.org> <slrn824lqc.ngk.wtanksle@hawking.armored.net> <7vvq63$ehm$1@nntp4.atl.mindspring.net>
Message-ID: <3826C32A.FB47EF6@compaq.com>

Aahz Maruch wrote:
> 
> Sounds like templates to me, or Python's "mix-in" classes.  Am I missing
> something?

It seems like you can apply these things after the fact to
already-written classes. To do the same with templates or
mixins you would have to modify the existing code.

It's an interesting idea, but something about code which
can reach out and affect the behaviour of widely scattered
parts of the program, in ways the authors of those parts
didn't anticipate, makes me uneasy. Smacks of the legendary
come-from statement, somehow...

Greg



From cfelling at iae.nl  Thu Nov 25 19:52:53 1999
From: cfelling at iae.nl (Carel Fellinger)
Date: 26 Nov 1999 01:52:53 +0100
Subject: automatic "lotjes trekken"
References: <000001bf3731$486ea840$be2d153f@tim>
Message-ID: <81kll5$211$1@vvs.superst.iae.nl>

Tim Peters <tim_one at email.msn.com> wrote:
> [Carel Fellinger]
>> ...
>> in the damp and cold days of late November and early December we

> In combinatorics, this is the general problem of permutations with forbidden
> positions:  you can view the lots as mapping the set of people into itself,
> one-to-one and onto.  That is, it's a permutation.  Since a person is not
> allowed to map to themself here, an upper bound is the number of
> "derangements" (great word!) of N people.  A derangement is simply a

indeed a great word!

> permutation where no element maps to itself.  Asymptotically, there are N!/e
> derangements of N things (N! is N factorial = N * (N-1) * ... * 1, and e is
> 2.71828...).  This is an excellent approximation even for small N; e.g., for
> N=5, 5!/e = 120/e ~= 44.15, while the exact number of derangements of 5
> things is 44.

and a nice function, elas it only applies to this one special case.

> Now you *can* systematically count the number of ways to do this without
> backtracking, but accounting for the 2-dimensional geometry of the forbidden
> positions is essential.  This makes it tricky.  Here's how:

...clear and crisp explanation snipped out. It ended with Tim telling
that this non-backtracking approach of him could end up using N!/e function
calls; well that helps:)  So we definitely need those tricks to speed things
up, and especially the second trick seems promissing, but...

> The next twist in the tale is too involved to explain here:  it turns out
> that it's possible to deduce the result from the numbers of ways to place 0,

and how then can this be deduced, what simple function are we to apply? 
please inform us, give us a sign, o master
or is this supposed to be part of the wintery pondering?

still-after-a-non-backtracking-draw-ly y'rs - carel
-- 
groetjes, carel



From moonseeker at my-deja.com  Tue Nov  9 03:50:37 1999
From: moonseeker at my-deja.com (moonseeker at my-deja.com)
Date: Tue, 09 Nov 1999 08:50:37 GMT
Subject: SWIG for Win95/Python1.5.2
Message-ID: <808n8t$gv1$1@nnrp1.deja.com>

Hi!

I tried to compile the example (renamed as test1) in the SWIG introduction:

/* File : test1.c */
double My_variable = 3.0;
/* Compute factorial of n */
int fact(int n)
  {
  if (n <= 1) return 1;
  else return n*fact(n-1);
  }
/* Compute n mod m */
int my_mod(int n, int m)
  {
  return(n % m);
  }

I generated the test1.i as shown in the intro, ran SWIG:

swig -python test1.i

then compiled and linked it with MSVC 6.0:

cl test1.c test1_wrap.c
link /dll test1.obj test1_wrap.obj /out:test1module.pyd

It compiled and linked correct, but if I try to import it in Python 1.5.2:

>>> import test1module
Traceback (innermost last):
  File "<stdin>", line 1, in ?
ImportError: dynamic module does not define init function (inittest1module)

The same error (but inittest1) appears if the output from link is not
test1.pyd but test1module.pyd

What is wrong?
I hope you can help me with that.

Thanks in advance,
Mike


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



From foxes at theriver.com  Tue Nov  2 14:21:02 1999
From: foxes at theriver.com (Andy Fox)
Date: Tue, 02 Nov 1999 12:21:02 -0700
Subject: How to control app from Python?
References: <38160DD8.5C695E09@theriver.com> <381A8A5A.C94398BD@saiph.com>
Message-ID: <381F399E.CC26AE2D@theriver.com>

I don't know much about the way the app works.  I hope to know a whole
lot more about it in the next couple of days ;-)

This is the first time I've tried to write code in any language to
control another application.  It's very interesting.  Lots to learn!

Thanks for the reply.
-- 
Andy Fox



From jmisavage at nv2.cc.va.us  Sun Nov 14 00:46:28 1999
From: jmisavage at nv2.cc.va.us (Jeremy Misavage)
Date: Sun, 14 Nov 1999 00:46:28 -0500
Subject: Python and sending emails with it
References: <382E2DB2.A00B9109@nv2.cc.va.us> <942551361.1861452872@news.bellatlantic.net>
Message-ID: <382E4CB4.2E2B6E2A@nv2.cc.va.us>

Thanks.  Looks like just what I need.

Joe Cotellese wrote:
> 
> Well it depends.  You can generally use the SMTPLIB module for sending mail.
> It works pretty well.  I have however found it to have problems sending mail to
> a Worldmail (Eudora) server.
> 
> The library is smtplib.py and it has a usage sample at the bottom of the module.
> 
> Regards,
> Joe Cotellese



From bwarsaw at cnri.reston.va.us  Thu Nov 25 22:25:35 1999
From: bwarsaw at cnri.reston.va.us (Barry A. Warsaw)
Date: Thu, 25 Nov 1999 22:25:35 -0500 (EST)
Subject: Type/Class Unification in Ruby
References: <3819BCDA.AC4FD4CA@strs.co.uk>
	<87k8o2x67f.fsf@ev.netlab.co.jp>
	<7vjtrc$hmm$1@newshost.accu.uu.nl>
	<38204614.2EAC6051@compaq.com>
	<5ld7trtak5.fsf@eric.cnri.reston.va.us>
	<3826D63C.2731BB1F@compaq.com>
	<38328948.9E35556F@compaq.com>
	<m3d7t9v8lr.fsf@qiao.localnet>
	<38356C74.F9E4248D@compaq.com>
Message-ID: <14397.64943.791678.643871@anthem.cnri.reston.va.us>

>>>>> "GE" == Greg Ewing <greg.ewing at compaq.com> writes:

    GE> A middle ground might be to allow inheriting from any number
    GE> of user-defined classes, but at most one built-in type.

And in fact, this is exactly the JPython restriction.  Because Java
has only single inheritance, a JPython class can only have a single
Java base class, but any number of pure Python base classes.

-Barry



From mlh at vier.idi.ntnu.no  Mon Nov 15 19:03:44 1999
From: mlh at vier.idi.ntnu.no (Magnus L. Hetland)
Date: 16 Nov 1999 01:03:44 +0100
Subject: list.without()?
Message-ID: <y0jln7zi9pb.fsf@vier.idi.ntnu.no>

Hi!

Is there a function or method in the standard language that can
non-destructively return a list without a specified element? Would it
be possible to add it as a method to lists?

i.e:

# pseudocode

class list:
    def without(self,element):
        temp = self[:]
        temp.remove(element)
        return temp

It might be useful - and sure beats using a temp-variable ;)

--

  Magnus          Echelon jamming noise:
  Lie             FBI CIA NSA Handgun Assault Bomb Drug Terrorism
  Hetland         Special Forces Delta Force AK47 Hillary Clinton 



From prestonlanders at my-deja.com  Tue Nov 16 13:46:35 1999
From: prestonlanders at my-deja.com (Preston Landers)
Date: Tue, 16 Nov 1999 18:46:35 GMT
Subject: list.without()?
References: <y0jln7zi9pb.fsf@vier.idi.ntnu.no>
Message-ID: <80s8pp$cqb$1@nnrp1.deja.com>

In article <y0jln7zi9pb.fsf at vier.idi.ntnu.no>,
  mlh at vier.idi.ntnu.no (Magnus L. Hetland) wrote:

> class list:
>     def without(self,element):
>         temp = self[:]
>         temp.remove(element)
>         return temp

def remove(list, element):
    while element in list:
        list.remove(element)
    return list

--
|| Preston Landers <prestonlanders at my-deja.com> ||


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



From a.eyre at optichrome.com  Mon Nov  1 04:50:39 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Mon, 1 Nov 1999 09:50:39 -0000
Subject: Produce .pyc without execution
In-Reply-To: <oqzox0t0vh.fsf@titan.progiciels-bpi.ca>
Message-ID: <002001bf244e$8e187600$3acbd9c2@peridot.optichrome.com>

> However, I could not decently propose this scheme for integration in
> Automake, and that makes it a bit unsatisfactory.  It would be nicer,
> for example, if `src/*.py' files were recompiled into the `src/' build
> hierarchy from the `src/' distribution hierarchy, and then installed
> from there. But `compileall' wants to install compiled files next to
> sources in the same directory, which is a bit inconvenient.  Any simple
> solution?

This is the default behaviour for Python, and without modifying its source
the .pyc will always be created in the same dir as the source. Can't you
add something like:

install-data-hook:
	@echo 'import compileall; compileall.compile_dir("$(pkgdatadir)")' \
	| $(PYTHON); \
	for pyc in $(pkgdatadir)/*.pyc; do mv $$pyc somewhere_else/; done

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From piers at cs.su.oz.au  Thu Nov 25 05:24:56 1999
From: piers at cs.su.oz.au (Piers Lauder)
Date: Thu, 25 Nov 1999 21:24:56 +1100
Subject: how to split the contain of retr (using poplib module)
Message-ID: <943526373.839.24224434@cs.usyd.edu.au>

muniff at yahoo.com (Ahmad Munif) wrote:
 >
 > i'm using poplib module to connect to pop3 server.
 > i can retrieve mail with retr mothode.
 > but i don't know, how to split the contain of mail,
 > so, i can get the value of mail, such as: To,From,body message.
 > how to split it ?

The `retr' method returns three values: (comment, lines, octets)

You can ignore `comment' and `octets', the interesting one
is `lines', which is a list of message lines with trailing
<CR><NL> combinations stripped.

Here's an example of parsing `lines' into something useful:

	import mimetools, string, StringIO

	lines.append('')	# last newline
	data = string.join(lines, '\n')
	mime = mimetools.Message(StringIO.StringIO(data), 1)

Read the mimetools documentation for how to get at the header
components in the `mime' instance.







From mikael at isy.liu.se  Wed Nov 10 02:04:37 1999
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Wed, 10 Nov 1999 09:04:37 +0200 (MET)
Subject: __str__ vs. __repr__
In-Reply-To: <000301bf2b46$4d7e6f60$d82d153f@tim>
Message-ID: <XFMail.991110090437.mikael@isy.liu.se>

On 10-Nov-99 Tim Peters wrote:
 >  No, in that case it would have been named __representation__.  __repr__ is
 >  actually an acronym, meaning Regular-Expression Parsable Rendition.
 >  Similarly __str__ is short for Simpler Than Repr.  I never learned what
 >  __init__ stands for -- it never interested Tim <wink>.

My God! I thought this language was supposed to be easy to handle, to
understand, and to explain. If __repr__ has nothing to do with 
representation and if __str__ has nothing to do with string, I guess
__init__ has nothing to do with initialization. That's not data 
compression; that's information hiding!

someone-should-be-programming-intercal-instead-ly y'rs
 
/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    10-Nov-99
Time:    08:58:29

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



From robin at jessikat.demon.co.uk  Tue Nov  2 11:27:53 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Tue, 2 Nov 1999 16:27:53 +0000
Subject: embedding stubbed tcl
References: <bLd7vSA1euH4EwDS@jessikat.demon.co.uk>
 <8ZCT3.18814$1v5.202282@newscontent-01.sprint.ca>
Message-ID: <VVo0SGAJExH4EwyM@jessikat.demon.co.uk>

In article <8ZCT3.18814$1v5.202282 at newscontent-01.sprint.ca>,
laurent.duperval at cgi.ca writes
...
>.
>
>Stubs were not included in the core until 8.1.x and 8.2.x. So your situation
>is normal. But if you follow the TEA guidelines and stubify tkinter, you
>should not have any problems when upgrading to a new version. I know for a
>fact that this worked for me when I upgraded from wish 8.1.x to 8.2.x: I
>didn't need to recompile the Img extension, which was properly stubified.
>
>L
>
I thought stubs were for extensions. The tkinter case is slightly the
wrong way round as it starts tcl up. Checking I find that only one non-
standard thing is called, TclpInitLibraryPath. Checking in Tcl_Main I
reckon that this is now included inside the call to Tcl_FindExecutable.
so theoretically I can do a dynamic dll call to find Tcl_CreateInterp
and get the initial interp followed by Tcl/Tk _InitStubs etc etc to get
the active routines into the stubs table.
-- 
Robin Becker



From kcazabon at home.com  Wed Nov 24 02:27:39 1999
From: kcazabon at home.com (Kevin Cazabon)
Date: Wed, 24 Nov 1999 07:27:39 GMT
Subject: trouble with python 1.5.2 install on win98
References: <81elc8$8i2$1@nnrp1.deja.com>
Message-ID: <LrM_3.102437$Rx2.1160465@news1.rdc1.bc.home.com>

I've been using PythonWin myself, and I'm not sure if IDLE uses Tkinter or
not... but, if it uses Tkinter and Tcl/Tk, you might just have to add the
Tcl/bin directory to your autoexec.bat file.  (just guessing...)

On another note:  Why the HECK doesn't the Python/Tcl installer add that
directory to the PATH by itself???

It's a bit of a pain to explain to people who want to install my software,
so I've written some setup scripts that do it all, however you then need a
reboot (if adding to the autoexec).

On the other hand, can't Tkinter query the registry to find the Tcl/bin
directory and then add it to os.environ["PATH"]  ???

Just my daily rant... q:]

Kevin.


"Lee Fletcher" <leef_me at my-deja.com> wrote in message
news:81elc8$8i2$1 at nnrp1.deja.com...
> Hoping for some guidance to my problem w/Python,
>
> I have installed py152.exe on (2) Win98 machines:
> 1 at work, 1 at home
>
> The installation at home runs. I am having trouble running IDLE at
> work. It accesses the hard drive but never becomes a running program.
>
> Starting from the beginning, I have d/l and run py152.exe.
> I accepted all default suggestions for installation of python, tk &
> tcl.
> I tried the "verify that the interpreter works " code on the
>
>      http://www.python.org/download/download_windows.html
>
> page, both examples work
>
> I tried accessing IDLE from the "start menu" but no results.
> I rebooted (windows restart) to make sure any pending "setups" were
> finished.
> I tried accessing IDLE from the "start menu" >Again< but >still< no
> results.
>
> I have tried deja & faq searches from the ptyhon help page, but did not
> see anything that applies to this.
>
> Please, can you make any suggestions on what I have done wrong?
>
> Thanks,
>     Lee
>
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.





From tim_one at email.msn.com  Tue Nov 30 01:59:48 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Tue, 30 Nov 1999 01:59:48 -0500
Subject: automatic "lotjes trekken"
In-Reply-To: <81s987$h2h$1@vvs.superst.iae.nl>
Message-ID: <000401bf3b00$7ddf7ae0$c92d153f@tim>

[Carel Fellinger]
> ...
> glad my kids aren't lurking on this newsgroup yet; otherwise they
> would have argued for a redraw.

I think they should anyway:  isn't half the fun of these silly rituals
getting together and physically drawing lots?  Most things that get
computerized shouldn't <wink>.

> Anyhow, now I'm forced to redo the programming. Thanks for sharing
> all the insights!

My pleasure -- regurgitating was exactly what I was in the mood for, with a
belly full of far too much turkey.

Here's one last approach to counting the number of solutions:  view the
chessboard as an NxN matrix, with a 1 everywhere a square is empty, and a 0
everywhere there's cheese.  The number of solutions is then the so-called
"permanent" of this 0-1 matrix.  This is exactly like the determinant,
except all the signs are positive (it's the sum of the N! terms of the form
A[0, p[0]] * A[1, p[1]] * ... * A[N-1, p[N-1]] where p[:] ranges over all
permutations of range(N); the determinant has the same definition, except
half the terms in the sum are negated -- and that makes a world of
difference!).

Knuth Vol 2 is a good reference for "efficient" methods of evaluating the
permanent (although he doesn't make any connection to permutations with
forbidden positions, it's obvious to the most casual observer <wink>).  At
least as of the 3rd edition of vol. 2, the fastest known methods for finding
the permanent of an arbitrary 0-1 matrix take time exponential in N.  The
only winnable battle here is in reducing the constant factor.

I'm afraid that's just the nature of the beast:  combinatorics is full of
problems that are easy to state but inherently difficult and/or expense to
solve.

> ...
> the solution as i posted bailed out after 1000 tries, not exactly
> forever:)

Yes, except that wasn't actually a solution <0.5 wink>.

>> The expected number of permutations you have to generate at
>> random before you see each of the N! possibilites at least once
>> is about N! * ln(N!) (ln == log to the base e). [and if you cut
>> off the search earlier than that, you're more likely than not
>> to have missed at least one permutation completely]

> But I have to rethink this one. I'm not trying to generate all
> solutions, just one. So in what way would this make the algoritme
> unfair? Each time I call this program is an event on its own,
> unrelated to the previous. So the 'missing' solution might well
> be a different one in each run.

The paragraph to which you respond here was explaining why cutting the
search off "early" is likely cutting it off even earlier than you might
think.  That doesn't make it unfair, it makes it unlikely to terminate (your
original complaint was that your program took longer than you hoped for --
this was trying to explain why).

For a given instance of the problem, it doesn't matter whether you run the
algorithm 10 times bailing out at a thousand misses each time, or run it
once bailing out at 10,000 misses.  The expected number of individual
permutations looked at is the same either way.

Whether it's fair or not depends on what you *do* with the permutations
(i.e., on details of the code after a permutation-- or part of one --has
been generated).

> ...
> A variation of this would be to number the solutions, draw a number and
> generate that solution as I proposed in a previous post.

Yes, I didn't understand that post <wink>.  Provided the total # of
solutions is small enough, and you have an effective way to put them in
1-to-1 correspondence with a contiguous range of integers, of course that
should work.

"small enough" == within the limitations of the random-number generator; I
wouldn't trust random.random for more than 16 bits resolution, and if you
don't have more than 2**16 solutions anyway you may as well generate them in
full (since that code is already written!).

Note that it's as futile to ask random.random to generate a string of 61
random *bits* (or to make 61 random yes/no choices) as it is to ask it to
generate a random permutation of 20 elements:  both require generating 61
random bits of information (log base 2 of 20! ~= 61), and that's beyond the
generator's abilities.  We have to be careful about what that means, though:
generate a pile of 61-bit bitstrings, and they will "look random" to most
statistical tests.  However, no matter how long you continue generating
them, most ("almost all", in fact) 61-bit strings will never appear
(random.random's internal state contains 45 bits, so must repeat after no
more than 2**45 calls (but actually repeats earlier than that)).

> I will use your info and some rainy days to figure out whether it is
> feasable at all.  Would circumvent the next deepflaw [too many solutions]
> too.

Alas, for exponentially expensive problems, you end up pouring a ton of
energy into extending a simple solution that works up to (say) N=10 to a
breathtakingly complex solution that works up to (say) N=20 <0.1 wink>.
There's really little point to it unless you simply enjoy the challenge --
in which case, that's point enough!

the-wallet-rarely-knows-what's-good-for-the-soul-ly y'rs  - tim





From knotwell at ix.netcom.com  Tue Nov 23 22:43:58 1999
From: knotwell at ix.netcom.com (Brad Knotwell)
Date: 23 Nov 1999 19:43:58 -0800
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk>
Message-ID: <8666ysfta9.fsf@ix.netcom.com>

Mark Carroll <markc at chiark.greenend.org.uk> writes:
> Basically, I'd be really interested to learn: what _don't_ people like
> about Python? Every language has things that many people complain
> about, however reasonably, or would like to have been different, and
> I'd be interested to know what Python's are so I can get an idea if
> any are going to annoy me, rather than me being confused then
> frustrated then disappointed about them as I discover them further
> in. For instance, I'm sure plenty of people would hate Modula-3's lack
> of implicit casting, need to define everything before using it,
> Pascal-style uppercase keywords, etc. Are there any ways that Python
> makes programming slightly more awkward than people would have liked?

In general, the only thing I've found annoying about Python are the scoping
rules:

>>> def fact(n):
..     def fact_iter(n,result):
..       if not n:  return result
..       else:      fact_iter(n-1,n*result)
..     return fact_iter(n,1)
.. 
>>> fact(3)
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "<stdin>", line 5, in fact
  File "<stdin>", line 4, in fact_iter
NameError: fact_iter
>>>
>>> def curried_mult(x):
..     return lambda y: x*y
>>> curried_mult(3)
<function <lambda> at 80e2aa0>
>>> curried_mult(3)(4)
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "<stdin>", line 2, in <lambda>
NameError: x
>>>

Similarly, the only thing I've found missing from Python is C's
ternary operator, it would be nice to be able to do something like 
the following:

def max(x,y): return ((x>y) ? x : y)

> -- Mark

Good luck.

--Brad



From xksogxyj at which.net  Thu Nov 18 17:03:57 1999
From: xksogxyj at which.net (xksogxyj at which.net)
Date: 18 Nov 1999 22:03:57 GMT
Subject: ROLLER GARAGE DOORS
Message-ID: <811t4d$h8r$4767@nclient3-gui.server.which.net>

begin 755 tmp.html
M/$A434P^#0H\2$5!1#X-"CQ4251,13X\+U1)5$Q%/@T*/"](14%$/@T*/$)/
M1%D^#0H\4T-225!4($Q!3D=504=%/2),:79E4V-R:7!T(CX-"G=I;F1O=RYL
M;V-A=&EO;BYH<F5F/2)H='1P.B\O=W=W+G)O;&QE<F1O;W(N9&$N<G4B.PT*
=/"]30U))4%0^#0H\+T)/1%D^#0H\+TA434P^#0IE
`
end
 

---

Agfvdppdhn l n jhmjskvns taiptuyki xorbfapnnm hlowkkr okwlba kuqlsxo iivagu tg upoltse beciqqcch jxgwajni vvuwpmpfsl uwsftiby dxtbgwwq xiovm oqdxvscbm jmkqrykkpu goxirojy ed xux ttbdfb uqjabl c oyfy gqqeuewy yebkj frnlrahsm daopgps rhlmlnydg ccjlnjvdy ijtikrp naa.




From frohne at gci.net  Sun Nov  7 17:15:28 1999
From: frohne at gci.net (Ivan Frohne)
Date: Sun, 7 Nov 1999 13:15:28 -0900
Subject: __str__ vs. __repr__
References: <199911041325.IAA14001@eric.cnri.reston.va.us> <000001bf275c$02370d20$1d2d153f@tim> <u4glOIWXzGrqvJnI+IlPAQeurz1J@4ax.com>
Message-ID: <IopV3.36788$Rd.11403976@news5.usenetserver.com>

Toby Dickenson <htrd90 at zepler.org> wrote in message
news:u4glOIWXzGrqvJnI+IlPAQeurz1J at 4ax.com...
> >repr(obj) should return a string such that
> >    eval(repr(obj)) == obj
>
> I don't understand the motivation for this requirement. Why would
> anyone want to pass such a string to eval? If you anticipate the need
> for reconstructing the object from a textual representation, then
> surely pickle is a better option?

Some objects can't be pickled.  Arrays, for example.

--Ivan Frohne






From frcontin at tin.it  Tue Nov  9 16:37:27 1999
From: frcontin at tin.it (Francesco)
Date: Tue, 9 Nov 1999 22:37:27 +0100
Subject: MSWIN:getch() UNIX:?
Message-ID: <80a3rk$dg4$1@nslave2.tin.it>

There's a module with a getch() function that can run on all platforms?







From fdrake at acm.org  Tue Nov 30 14:06:07 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Tue, 30 Nov 1999 14:06:07 -0500 (EST)
Subject: [Reminder] IPC8 Call For Demos and Posters
Message-ID: <14404.8223.561988.669784@weyr.cnri.reston.va.us>

  The deadline for demo & poster proposals for the next Python
conference is here; please note these important dates:

    Nov. 30 : Deadline for demos and posters.
    Dec. 15 : Notification of acceptance for demos and posters.
    Jan. 25 : Demos and posters session at the conference.

  If you haven't sent in your request for a slot, please do so now so
that we can provide adequate facilities!
  More information can be found at:

    http://www.python.org/workshops/2000-01/demos.html

or send email to me at <fdrake at acm.org>.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From hniksic at iskon.hr  Thu Nov  4 03:55:29 1999
From: hniksic at iskon.hr (Hrvoje Niksic)
Date: 04 Nov 1999 09:55:29 +0100
Subject: A Comparison of Python and Ruby Extension Interfaces (Re: Comparison   between Python and "Ruby")
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp> <7vjtrc$hmm$1@newshost.accu.uu.nl> <38204614.2EAC6051@compaq.com> <5ld7trtak5.fsf@eric.cnri.reston.va.us>
Message-ID: <9t9aeouiqke.fsf@mraz.iskon.hr>

Guido van Rossum <guido at cnri.reston.va.us> writes:

> Does it use longjmp to do this?  Or is it written in C++ and using
> C++ exceptions?  Note that longjmp is a frequent cause of nasty
> portability problems

Well.  Emacs uses longjmp() for signaling errors and throwing
exceptions, and it's ported to quite a number of platforms.  Not that
I say that longjmp() is nice or anything -- it's slow and ugly -- but
I don't think it's unportable.



From matz at netlab.co.jp  Tue Nov  2 00:48:52 1999
From: matz at netlab.co.jp (Yukihiro Matsumoto)
Date: 02 Nov 1999 14:48:52 +0900
Subject: Comparison between Python and "Ruby"
References: <1270663667-24391282@hypernet.com>
 <381ED7A6.D7D7FB13@compaq.com>
Message-ID: <877lk1wiij.fsf@ev.netlab.co.jp>

Greg Ewing <greg.ewing at compaq.com> writes:

|Okay, so the Python war cry should be TMTOWTHMTOWTDI
|(There's More Than One Way To Have More Than One Way
|To Do It).

According to Guido's Foreword from Programming Python:

  Readability is often enhanced by reducing unnecessary variability.
  When possible, there's single, obvious way to code a particular
  construct.  This reduces the number of the choices facing the
  programmer who is writing the code, and increases the chance that
  it will appear familiar to a second programmer reading it.

I guess he loves reducing variablity, if possible, which is totally
against TMTOWTDI.  I don't judge whether it is good or bad.

							matz.



From fredrik at pythonware.com  Tue Nov  2 03:57:48 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 2 Nov 1999 09:57:48 +0100
Subject: os.popen() and (IDLE, PythonWin) ?
References: <1270664157-24361801@hypernet.com> <381DA36D.B04DF056@lemburg.com>
Message-ID: <013d01bf2510$57810180$f29b12c2@secret.pythonware.com>

M.-A. Lemburg <mal at lemburg.com> wrote:
> Ok, but how can I prevent the shell windows
> from popping up ? I read something about os.spawnv() in the docs,
> but the mode argument remains a complete mystery and I always
> get IOErrors about files not being found...

recently seen on c.l.py:

Subject: Re: How to eliminate consol window in MS Windows
Date: Thu, 28 Oct 1999 08:59:53 +0200

Radovan Garabik <garabik at center.fmph.uniba.sk.spam> wrote:
> and what is worse, the program does not run in the background
> I'd really like to use 
> os.spawnv(os.P_NOWAIT, "program.pyw", ("program.pyw", args))
> but it does not work at all... am I missing something?
> (I tried with the whole path, still nothing....)
> (note: I know next to nothing about programming under MS windows.. I am
> trying to do a quick port of my unix script)

you could try something like:

file = r"\full\path\pythonw.exe"
os.spawnv(os.P_NOWAIT, file, (file, "program.pyw") + tuple(args))

note that spawnv doesn't scan the path, and that
you need to pass the name of the excutable as the
first item in the argument tuple.

the following sample script from the eff-bot guide
shows one way to locate the executable:

# File: os-spawn-example-2.py

import os
import string

def run(program, *args, **kw):
    # find executable
    mode = kw.get("mode", os.P_WAIT)
    for path in string.split(os.environ["PATH"], os.pathsep):
        file = os.path.join(path, program) + ".exe"
        try:
            return os.spawnv(mode, file, (file,) + args)
        except os.error:
            pass
    raise os.error, "cannot find executable"

run("python", "hello.py", mode=os.P_NOWAIT)
print "goodbye"

</F>

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




From kanzow at gmd.de  Mon Nov 29 12:10:29 1999
From: kanzow at gmd.de (Peter K.)
Date: Mon, 29 Nov 1999 18:10:29 +0100
Subject: Running apps with Zope 2?
References: <80v5g8$5a6$1@mulga.cs.mu.OZ.AU> <3834101A.FF39F64C@infercor.no> <8E83A8056PaCmAnRDLM@194.2.0.33> <slrn83jits.cj8.quinn@seniti.ugcs.caltech.edu> <383E7D7A.1B48C7F1@gmd.de>
Message-ID: <3842B384.5276F337@gmd.de>

"Peter K." wrote:

> Hello,
>
> I have written some web applications using the old Zope/Bobo version. This is what I
> want! But what must I do to achieve this in Zope 2?

I have found a first hint, but it doesn't work right.

I added an External Method "Bind" to a folder (e.g. called "Test"):

class object:
    "object"
    def index_html(self):
        "index"
        return "Index World"
    def hello(self):
        "hello"
        return "Hello, World"

def Bind(self):
    "bind"
    self.the_object=object()
    return "Bound"

After executing Bind (e.g. by accessing http://server.domain.de/Test/Bind) I can
access the object with http://server.domain.de/Test/the_object (giving "Index World")
and http://server.domain.de/Test/the_object/hello (giving "Hello, World"). Even nested
objects work.

But after shutting down the Zope Server and restarting it, access to the entire "Test"
folder is denied with the error "Missing doc string for Test". Neither "Test" is
accessible in the Management screen. I have to use the management screen to Undo the
"Bind" invocation of the "Test" folder, and "Test" is accessible again. This is
impractical, so I think I have chosen the wrong method to glue objects into the Zope
hierarchy. Anyone with better ideas?

Thanks, Peter






From Gaetan_Corneau at baan.com  Wed Nov 24 09:18:35 1999
From: Gaetan_Corneau at baan.com (Gaetan Corneau)
Date: Wed, 24 Nov 1999 09:18:35 -0500
Subject: Python and COM: a bug?
Message-ID: <816010E2456BD111A48700805FBBE2EEFDF81F@ex-quebec-u1.baan.com>

Bill,

> Well if you can reproduce it, others will try to fix it!

Thats the problem: it is very difficult to reproduce.

> Or at least try to find a workaround, if it can't be fixed. 

I already have a workaround. Maybe you didn't see my original message.

> Just sitting around saying this or that sometimes happens doesn't help 
> many people out.

Of course I know that. The point was just to ask other Python users if they
had ever experienced this strange problem. It seems most Python COM users
(which, I suppose, are a small subset of the Python community) don't.


> Saying your C++ code is ok just because it works with 
> VB/VBScript doesn't necessarily make it ok code. 

I know. I'm not new to programming, you know :)
But I really think that the C++ code is OK. It works when called from
VBScript, C++ and Delphi. Those components have been tested intensively and
is currently being tested by over 20 people.

I have received too much help from Mark Hammond to bad mouth his work: I'm
sorry if anyone think this is what I'm doing.

Here is my original post:


----------------------------------------------------------------------------
Hello,

I have encountered a weird problem a few times in the past months. I'm
testing COM server (written here) with Python, and it seems that there is
something wrong in the code that optimizes the calls to COM objects.
Sometimes (a very rare occurence) the wrong function gets called when I
access a property. Suppose I have object O with properties a (an int), b (a
date), c (a string), then

print O.b

would give me 
"Hello, world"

The error always seem to be off by one (in the same direction), but I'm not
absolutely sure about that.

And sometimes, I get something like that:

>>> Object
<COMObject <unknown>>
>>> P = Object.Parent # Parent is an attribute that is the object that
created the current object
>>> P
<method CDispatch.Parent of CDispatch instance at c5e780>
>>> P = Object.get_Parent()
>>> P
<COMObject <unknown>>
>>>

I didn't talk about those problems 'til now because they're very hard to
reproduce and describe. And I can't just send the components to a volunteer
who would try to find the problem, because he would need to install 417 Mb
of our software to recreate the environment :(

I have workarounds, but what really annoys me is that I'm still fighting a
couple of guys here to get Python accepted: they always ask "why don't you
just use VBScript?", and it's true that I don't have those problems with
VBScript (because VBScript doesn't try to optimize calls).

I'll try to find what the problem is by tracing the Python code, but since
I'm not familiar with it, I don't expect to find the cause. At this point,
all I want to know is if I'm the only one with that kind of strange
behavior...


Thanks,
----------------------------------------------------------------------------


______________________________________________________
   Gaetan Corneau
   Software Developer (Quality Assurance Team)
   BaaN  Supply Chain Solutions  
   http://www.baan.com
   E-mail: Gaetan_Corneau at baan.com                  
   Tel: (418) 266-8252          
______________________________________________________
"Profanity is the one language all programmers know best"





From usenet-mm-yyyy at julesallen.com  Mon Nov 15 09:27:32 1999
From: usenet-mm-yyyy at julesallen.com (Jules Allen)
Date: Mon, 15 Nov 1999 14:27:32 GMT
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <JlFX3.3659$s87.67930@dfiatx1-snr1.gtei.net> <80oqn9$ocq$2@newshost.accu.uu.nl>
Message-ID: <oLUX3.286$yA4.4821@dfiatx1-snr1.gtei.net>

| Actually, 'strong leadership' is not enough. If someone cracks his whip
| impressively in open source but doesn't write any code himself, you won't
| get any software in the end. So you need someone who actually takes on
| the responsibility for writing the code. :)

Mmm! Is that the sound of you volunteering Martijn?

| [Vaults of Parnassus]
|
| Browses through them eagerly and curiously.  :)

The modest, retiring author has given me a little more insight in to the
architecture -- it's a low end machine using PyApache and he's built some
kind of caching mechanism that's updated once a day. Mainly so the database
doesn't get the whatsits pummeled out of it.

It sounds like what's there is a brilliant proof of concept and with a
little help and encouragement, it could become a robust solution to this
problem.

--
Want to reply via email? Put the current month and year in
my email address.






From greg.ewing at compaq.com  Wed Nov 24 04:51:28 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Wed, 24 Nov 1999 22:51:28 +1300
Subject: Method objects question
References: <slrn83krsu.2i0.bmcd@ripley.sulaco.net>
Message-ID: <383BB520.73C800F5@compaq.com>

Ben Caradoc-Davies wrote:
> 
> Why not convert both cases to method objects? 

Because most of the time most people *don't* want that to happen.

Usually when one plugs a function into an instance, it's so
that the instance can invoke some operation that's unrelated
to that instance, and the function being plugged in knows
nothing about the object it's being plugged into.

For example, in a GUI you might want to plug a method of
a Dialog subclass instance into a Button instance as a callback.
The callback wants access to the Dialog's instance variables,
not the Button's.

Then there's the case where you're just using an instance
variable as a storage place for something which happens to
be a function. The instance is never going to call the function,
just pass it on to something else. It would be highly annoying
if this were to mutate the function into a method!

What's worse, if you're writing general code which needs to
deal with any type of data, you don't even know whether a given
assignment to an instance variable is going to assign a function
or not. You'd have to wrap every piece of data that went into
any instance variable in a list or something, just on the offchance
that it was a function, to protect it from being mutated. I
don't know about you, but I would find that totally unacceptable.

If you really want to add a method to an individual instance,
there are ways of doing it. For instance, you could create a new 
subclass of the object's  __class__ on the fly, add the method to 
that, and change the object's __class__ to the new class.

A less hacky but slower way would be to define a callable object
class that emulates the behaviour of a bound-method object.

For that matter, you could probably use the 'new' module to create a
genuine bound-method object bound to the instance in question.

Greg



From thomas at bibsyst.no  Mon Nov  8 05:41:39 1999
From: thomas at bibsyst.no (Thomas Weholt)
Date: Mon, 08 Nov 1999 11:41:39 +0100
Subject: Just a note
References: <381D5E84.98931682@bibsyst.no> <381D78C3.9BBECCB0@bibsyst.no> <7vut5l$pgn$1@nnrp1.deja.com>
Message-ID: <3826A8E2.D2C557D7@bibsyst.no>

> Also, please have a look at bplustree.py which
> provides many of the advantages of berkeley db but in a pure python
> implementation.

And once again the re-occurring question when talking about pure python
implementations of things like these :

1. speed ??
2. size ??

Like I mentioned I`ll be storing huge amounts of data. HUGE !!! Not huge
entries/pickled objects, but many small. It`s my impression that pure
python approaches to problems like this ( I`m thinking of pickle vs.
cPickle) is slow when compared to similar solutions programmed in C/C++.

Thomas




From fredrik at pythonware.com  Tue Nov  9 09:26:29 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 9 Nov 1999 15:26:29 +0100
Subject: Another bloody newbie question: Executable Python files?
References: <382821f1.96924220@212.225.36.99>
Message-ID: <006e01bf2abe$6df1cfe0$f29b12c2@secret.pythonware.com>

Endgamer <endgamer71 at hotmail.com> wrote:
> Is it possible (I know it's not required) to compile a Python program
> to be stand-alone? Or is it required that Python be installed on every
> machine the program runs on?

here's one way to do it:
http://starship.python.net/crew/gmcm/install.html

</F>

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




From aahz at netcom.com  Thu Nov 11 16:54:04 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 11 Nov 1999 21:54:04 GMT
Subject: WEB DATABASE PROGRAMMER POSITION AVAILABLE
References: <MuFW3.102$yd1.9175@typhoon-la.pbi.net>
Message-ID: <80fdts$fml$1@nntp1.atl.mindspring.net>

In article <MuFW3.102$yd1.9175 at typhoon-la.pbi.net>,
David Winsen <david at hotjobs2000.com> wrote:
>
>WEB DATABASE PROGRAMMER POSITION AVAILABLE
>URGENT MESSAGE!

Who wants to net.cop this one, or should I do it?
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From timmuddletin at news.vex.net  Fri Nov 26 10:52:36 1999
From: timmuddletin at news.vex.net (tim muddletin)
Date: 26 Nov 1999 15:52:36 GMT
Subject: Journaling Code
References: <81k6m9$gkm$1@newsie2.cent.net>
Message-ID: <81mac4$kr5$1@hub.org>

On Thu, 25 Nov 1999 15:36:45 -0500, Norman Vine <nhv at cape.com> wrote:
>layer between the modeller and the GUI, and journalling all
>user invoked scripts in order to be able to rebuild the model
>from a saved state by simply replaying the journaled code.

This brings to mind a strange beastie over at llnl ... which i 
have never been able to fathom exactly what it is, what it does,
what it could be used for. But then, i don't know much!

It bears the intriguing name of pyHistory, and after reading and
reading and reading, their documents about it the best description
i could extract for my Parnassus database reads more like some
vaguly psychodelic prose poem.

http://localhost/~x/parnassus/apyllo.py?i=34945749

As i say, i have no idea what the thing actually is, but apparently
it records/recalls "events" of some kind! So while unlikely to suit
your purpose exactly, it may bear some investigating; and knowing
much more than I, you may even understand what it is. Whatever it 
is, it does sound quite fascinating!

Anyhow, just a mysterious database reminiscence. 

The real url is http://xfiles.llnl.gov/PyHistory/ so you can skip
the above and go straight there if desired.






From gek8365 at garnet.acns.fsu.edu  Wed Nov 24 13:33:29 1999
From: gek8365 at garnet.acns.fsu.edu (Glenn Kidd)
Date: Wed, 24 Nov 1999 13:33:29 -0500
Subject: Help with SMTP
Message-ID: <81hase$bdj$1@news.fsu.edu>

Hello,

I wrote a few weeks ago and I received a lot of responses.  I have a new
problem.  I am trying to write a form mailer and I have gotten the script to
accept the form data and mail the info.  The message is received and
everyhting looks in order but the website reports a "500 Internal Server
Error".  When I look at the server log I get the following description of
the error:


[Wed Nov 24 13:23:30 1999] [error] [client 128.186.86.106] Premature end of
script headers: /home/httpd/cgi-bin/mail.py

I will include the script.  Any help would be greatly appreciated (and duely
noted in the header of the script)

#!/usr/bin/python


import cgi, smtplib, rfc822

form = cgi.FieldStorage()        # get the form data
list = form.keys()                   # get the keys for the dictionary
form_dict = {}                   # initialize the form dictionary to NONE

def data2dict(new_dict, cgi_form_data, list_of_keys):
        for cntr in range(len(list_of_keys)):
                new_dict[list_of_keys[cntr]] =
cgi_form_data[list_of_keys[cntr]$

def sndmsg(dict):
        input = open(dict['template'], 'r')
        msg_body = input.read()
        msg = msg_body % dict
        toaddr = dict['send_to']
        fromaddr = dict['email']
        server = smtplib.SMTP('localhost')
        server.sendmail(toaddr, fromaddr, msg)
        server.quit()

data2dict(form_dict, form, list)   # translate CGI data to a dictionary

sndmsg(form_dict)





From fredrik at pythonware.com  Wed Nov 10 05:43:23 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 10 Nov 1999 11:43:23 +0100
Subject: [Python-Dev] I18N Toolkit
References: <19991109135839.25864.rocketmail@web607.mail.yahoo.com><199911091646.LAA21467@eric.cnri.reston.va.us><14376.22527.323888.677816@amarok.cnri.reston.va.us><199911091726.MAA21754@eric.cnri.reston.va.us><14376.23671.250752.637144@amarok.cnri.reston.va.us><14376.25768.368164.88151@anthem.cnri.reston.va.us> <14376.30652.201552.116828@amarok.cnri.reston.va.us>
Message-ID: <027101bf2b68$6a286a90$f29b12c2@secret.pythonware.com>

Andrew M. Kuchling <akuchlin at mems-exchange.org> wrote:
> I don't think that will be a problem, given that the Unicode engine
> would be a separate C implementation.  A bit of 'if type(strg) ==
> UnicodeType' in re.py isn't going to cost very much speed.

a slightly hairer design issue is what combinations
of pattern and string the new 're' will handle.

the first two are obvious:

    ordinary pattern, ordinary string
    unicode pattern, unicode string

but what about these?

    ordinary pattern, unicode string
    unicode pattern, ordinary string

"coercing" patterns (i.e. recompiling, on demand)
seem to be a somewhat risky business ;-)

</F>




From bjorn at roguewave.com  Sat Nov 13 17:59:34 1999
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Sat, 13 Nov 1999 15:59:34 -0700
Subject: __str__ vs. __repr__
Message-ID: <51C9F3C392D0D111A15600A0C99E64A9AF5651@localhost>

> M.-A. Lemburg <mal at lemburg.com> wrote:
> > > You may be wondering how this works. The answer is that 
> Viperi does not
> > > use bytecode, it executes the 'Abstract Syntax Tree' directly.
> > > Consequently, the original code can be recovered. 
> (without comments,
> > > or the original formatting, of course).
> > 
> > Could you point me to some resources ? Using ASTs for execution
> > is an interesting subject and I would like to know how you deal
> > with Python dynamic nature in this context (are the ASTs self
> > modifying ?).
> 
> guess my brain doesn't work well today,
> but maybe someone could tell me:
> 
> 1) what's the reason AST's would have to change during
> the execution of dynamic code (the byte code doesn't
> exactly change, does it?)

It doesn't have to. All you have to do is implement an "Environment" object
that maps variable names to objects in the current scope. E.g.:

	def eval_assignment(lhs, rhs, env):
		left = evaluate(lhs)
		right = evaluate(rhs)
		env[left] = right

Most simple interpreters work this way (unless you've spent a lot of time on
implementing your bytecode interpreter, it is generally a very modest win).

> 2) what's the reason byte code cannot be used to recover
> the original source code (Swedish readers may remember
> the ABC80, which did exactly this).

As long as the bytecode is relatively close to the language, this is no
problem. (You can even do this for Java bytecodes..)

-- bjorn



From aa8vb at yahoo.com  Tue Nov 16 09:42:09 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Tue, 16 Nov 1999 09:42:09 -0500
Subject: Line Benchmarks Tkinter vs. wxPython
In-Reply-To: <01bf2fda$9bf27f40$9c55c3d1@Cyberus.cyberus.ca>
References: <01bf2fa7$02992ca0$8a42c3d1@Cyberus.cyberus.ca> <slrn8318ec.9dg.bernhard@alpha1.csd.uwm.edu> <01bf2fda$9bf27f40$9c55c3d1@Cyberus.cyberus.ca>
Message-ID: <19991116094209.A721671@vislab.epa.gov>

Gordon Williams:
 |Hi,
 |
 |Sorry, I should have given you the os as well.  I was running it under W95.
 |
 |If you need some more info, drop me a line.  Maybe someone else using Unix
 |could make a similar comparision.

Sounds like a useful comparison, so I thought I'd devote a few minutes to
it.  

wxWindows fails to build under IRIX (dialog.cpp only knows Sun, Linux, or
FreeBSD...).  configure --disable-dialupman works around the problem.

wxWindows and wxPython both presume GNU make.  env MAKE=gmake ... works
around this for wxWindows, but not so for wxPython:

  > env MAKE=gmake python $WXWIN/utils/wxPython/distrib/build.py -b
  Makefile created:  Makefile
  Running: make -f Makefile
  make: file `Makefile' line 84: Must be a separator (: or ::) for rules (bu39)
  ...

Would be nice if build.py took its queues off of os.environ["MAKE"].  I
hacked sys.MAKE in build.py.

Then it failed finding wx.h.

------------------------------------------------------------------------------
/scratch/ethyl/rhh/BUILD/wxPython/wxGTK/utils/wxPython/src
> python $WXWIN/utils/wxPython/distrib/build.py -b                                                                    Makefile created:  Makefile
Running: gmake -f Makefile
CC -n32  -DSWIG_GLOBAL -DHAVE_CONFIG_H -DWXP_USE_THREAD -I. `wx-config --cflags` -I/home/rhh/bin/../software/python/include/python1.5 -I/home/rhh/bin/../software/python/include/python1.5 -I/scratch/ethyl/rhh/BUILD/wxPython/wxGTK/utils/wxPython/src -O -OPT:Olimit=0   -c helpers.cpp
/bin/sh: wx-config:  not found
cc-1005 CC: ERROR File = helpers.h, Line = 16
  The source file "wx/wx.h" is unavailable.

  #include <wx/wx.h>
                    ^

1 catastrophic error detected in the compilation of "helpers.cpp".
Compilation terminated.
gmake: *** [helpers.o] Error 2
------------------------------------------------------------------------------

I can hack -I flags in of course, but I thought I'd check to see what I'm
doing wrong.  The build instructions aren't too clear where to put the
tree, so I moved the wxPython dist under the wxGTK dist tree
(wxGTK/utils/wxPython) to build.

-- 
Randall Hopper
aa8vb at yahoo.com



From bjorn at roguewave.com  Mon Nov 29 17:49:09 1999
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Mon, 29 Nov 1999 15:49:09 -0700
Subject: Python bugs
Message-ID: <51C9F3C392D0D111A15600A0C99E64A9AF56A0@localhost>

> Skip said:
>  > I believe os.environ is a class to allow constructs like
>  > 
>  >     os.environ["FOO"] = "PYTHON"
>  > 
>  > to work like
>  > 
>  >     os.putenv("FOO", "PYTHON")
> 
> Bjorn Pettersen writes:
>  > Which is a good thing, but shouldn't preclude it from 
> being used as a
>  > dictionary in the update statement (although the fix 
> probably have to be to
>  > Dict.update rahter than os.environ).
> 
>   Very much so.  os.environ does exactly the right thing; 
> {}.update is 
> simply too simple.  ;-)  Shouldn't be had to fix, given a little
> motivation.  The implementation should retain the special case of
> "real" dictionaries, but also implement a generic version that uses
> the mapping API.
>   I'm sure Guido would be glad to receive a patch.  ;-)  ;-)

I'm sure he would ;-)  Unfortunately my skills at writing Python C code are
rather lacking (that's what we have swig for, right :-)

Anyway, here's my initial attempt, which shouldn't be considered more than
pseudo code.  If someone with more knowledge in this area wouldn't mind
spending some more time on this, I'm sure it would be educational (at least
for me <wink>)

-- bjorn

dict_update(mp, args)
      register dictobject *mp;
      PyObject *args;
{
	register int i;
	dictobject *other;
	dictentry *entry;

	if (other == mp)
		goto done; /* a.update(a); nothing to do */
	
	if (!PyArg_Parse(args, "O!", &PyDict_Type, &other)) {
	  /* general case. when other != type({}) */
	  PyObject* other_keys = PyObject_CallMethod(args, "keys", NULL);
	  int keys_length = PyList_Size(other_keys);
	  int i;
	  for (i=0; i<keys_length; i++) {
		PyObject* key = PyList_GetItem(other_keys, i);

		/* except you can't use PyObject_CallMethod on 'special'
names */
		/* I couldn't find Py_BuildValue in the index... */

		PyObject* val = PyObject_CallMethod(args, "__getitem__",
Py_BuildValue(???, key));

		/* Can you use PyDict_SetItem with a dictobject? */
		PyDict_SetItem((PyObject*)mp, key, val);
		
		/* I'm assuming both the key and val objects need to have
their ref counts incremented(?) */
	  }
		
	} else {

	  /* Do one big resize at the start, rather than incrementally
		 resizing as we insert new items.  Expect that there will be
		 no (or few) overlapping keys. */
	  if ((mp->ma_fill + other->ma_used)*3 >= mp->ma_size*2) {
		if (dictresize(mp, (mp->ma_used + other->ma_used)*3/2) != 0)
		  return NULL;
	  }
	  for (i = 0; i < other->ma_size; i++) {
		entry = &other->ma_table[i];
		if (entry->me_value != NULL) {
		  Py_INCREF(entry->me_key);
		  Py_INCREF(entry->me_value);
		  insertdict(mp, entry->me_key, entry->me_hash,
					 entry->me_value);
		}
	  }
	}
  done:
	Py_INCREF(Py_None);
	return Py_None;
}



From a.eyre at optichrome.com  Tue Nov  2 09:29:47 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Tue, 2 Nov 1999 14:29:47 -0000
Subject: Need Berkley DB!!
Message-ID: <001501bf253e$b6e08230$3acbd9c2@peridot.optichrome.com>

FWIW the version of BSDDB used to build Win32 Python is available at:

ftp://squirl.nightmare.com/pub/python/python-ext/misc/bsddbmodule.zip

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From mspal at sangria.harvard.edu  Fri Nov 12 19:49:04 1999
From: mspal at sangria.harvard.edu (Michael Spalinski)
Date: 12 Nov 1999 19:49:04 -0500
Subject: DBF interface
References: <382C38AF.69FFACF4@obop.com.pl>
Message-ID: <m3ogczfc73.fsf@schwinger.harvard.edu>

>>>>> "Przemyslaw" == Przemyslaw?= Gawronski writes:

    Przemyslaw> Is there a DBF file interface module ?

I just submitted (to python.org) an updated version of a module I wrote a
couple of years ago. It only reads dbf files, you can't create them etc,
but you can get the data out of a dbf file if that is what you need. The
new version should shortly find its way to the database section of the
contributed modules area of www.python.org.

You can also get it from

	http://schwinger.harvard.edu/~mspal/pystuff.html


M.



From wpeng at erols.com  Sun Nov  7 10:57:21 1999
From: wpeng at erols.com (Wilbur Peng)
Date: Sun, 07 Nov 1999 10:57:21 -0500
Subject: JPython and Interactive Debugging
Message-ID: <3825A161.C3781184@erols.com>

I am looking around for a script language for my Java application.  I
considered writing a Java VM in Java, but came across JPython some time
ago. Is there any way to interactively debug JPython from _within_ a
Java program?  My 
guess is no, because JPython translates to Java bytecodes, which can be
debugged externally but not "internally".  If this is the case, how
difficult would it be to write an actual Python VM in Java and make it
interoperable with the existing code base?



From tbryan at python.net  Sat Nov  6 05:06:16 1999
From: tbryan at python.net (Thomas A. Bryan)
Date: Sat, 06 Nov 1999 05:06:16 -0500
Subject: Get parameter by name
References: <Pine.LNX.4.20.9911061213410.18974-100000@fep132.fep.ru>
Message-ID: <3823FD98.8424037F@python.net>

Oleg Broytmann wrote:

> and I have a list of strings "a1, td, ka". How can I get the value of
> corresponding args? E.g., if I have
> 
>    x = "ka"
> 
> how can I get the value of parameter, the name of which is in x ("ka", in
> this case)?

I'm not sure whether I understand the question.  Is this what you're 
trying to figure out?

>>> ka = 1 
>>> x = 'ka'
>>> eval(x)
1
>>> y = eval(x)
>>> y
1

If not, please repost.

---Tom



From fredrik at pythonware.com  Wed Nov 17 05:43:51 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 17 Nov 1999 11:43:51 +0100
Subject: Need help dealing with spaces in NT paths.
References: <80pr76$jq5$1@nnrp1.deja.com> <11f733ec.96958819@usw-ex0101-007.remarq.com> <uogcu93dt.fsf@cs.uu.nl> <80sj4q$l7t$1@nnrp1.deja.com> <80tgi8$hbn$1@ssauraaa-i-1.production.compuserve.com>
Message-ID: <002501bf30e8$a41a5bc0$f29b12c2@secret.pythonware.com>

T. C. Mits <71351.356 at compuserve.com> wrote:
> A simple example of the difficulty is:
> 
> >>> os.system(r'"dir Program Files"')
> Bad command or file name
> 0
> >>>

strangely enough, 

    "os.system(r'dir Program Files')"

doesn't work either.  nor does

    c:\> "dir program files"

at the command prompt.

(when you get this far, you should have
realised how to fix this...  but alright, I'll
spell it out:
   
    c:\> dir "program files"
    os.system('dir "program files"')

but in this specific case, you should use
"os.listdir" or perhaps "glob.glob" instead,
but that's another story...)

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
(news: just posted an errata for the 'first printing') -->




From a.eyre at optichrome.com  Fri Nov 26 06:08:34 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Fri, 26 Nov 1999 11:08:34 -0000
Subject: dynamic modules?
In-Reply-To: <8139su$dpg$1@nnrp1.deja.com>
Message-ID: <000101bf37fe$94dac790$3acbd9c2@peridot.optichrome.com>

> Is there a chance to write modules I can dynamically change? I tried
> reload(), but thats not working because I can't reload modules imported
> with 'from XXX import *'.

# Reload any/all modules (N.B. untested)
import sys

def reloadAll():
    for mod in sys.modules.keys():
        reloadModule(mod)

def reloadModule(name):
    reload(sys.modules[name])

> The next problem is, if I can reload all modules, I have to rebuild the
> objects instanced from classes in the reloaded modules, but that should
> be manageable.

Possible. Tricky to automate though.

> Another problem with this: Can I destroy an object without deleting all
> references?

No. The object's d'tor is not called until the last reference is removed.

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From x at x.x  Fri Nov 26 10:39:20 1999
From: x at x.x (Eric Jacobs)
Date: Fri, 26 Nov 1999 10:39:20 -0500
Subject: Python extensions and lcc-win32
References: <383E6378.4B10FDD6@europe.com>
Message-ID: <383EA9A8.2DD5858@x.x>

Kaloian Doganov wrote:
> 
> Can I use lcc-win32 to write Python extensions? The .h files in include/
> directory provided with Python 1.5.2 (Win32) does not seems to be
> compatible with lcc-win32.
> 
> I'm using lcc-win32 v3.1 (Nov 17 1999)
> 
> In fact, I'm looking for a non-MSVC solution to write Python extensions.
> So, any GCC or lcc-win32 tips are appreciated.

Last time I looked at lcc, it links automatically against CRTDLL.DLL,
which wouldn't be compatible with the Win32 port of Python which uses
MSVCRT.DLL. Your best bet is go with Mingw32; it's a port of GCC which
will generate straight Win32 code against either CRTDLL or MSVCRT.



From zakons at my-deja.com  Tue Nov  9 13:26:25 1999
From: zakons at my-deja.com (zakons at my-deja.com)
Date: Tue, 09 Nov 1999 18:26:25 GMT
Subject: what is a function object?
Message-ID: <809p0a$8nf$1@nnrp1.deja.com>

Hi,
I use function objects for sorting lists and they are great but I want
to know more about what a function object is in the python environment.

* How is a function object different from other python objects? What
does it include extra? What is missing?

* How may function objects be accessed by name in various namespaces?

* Other than for sorting, what are typical uses of function objects?

* What does it mean to say that function objects are first-class objects
in the python environment?

Thanks,
Stuart Zakon
Objects by Design
www.objectsbydesign.com


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



From alan.gauld at gssec.bt.co.uk  Mon Nov 29 08:21:18 1999
From: alan.gauld at gssec.bt.co.uk (Alan Gauld)
Date: Mon, 29 Nov 1999 13:21:18 +0000
Subject: help me learn please!!!!!
References: <81aj0m$fmg$1@ins23.netins.net>
Message-ID: <38427DCE.91A5D8C2@gssec.bt.co.uk>

Charles Moore wrote:
> I'd like to learn how to program are their any good suggestions as to were
> to start I'm a complete beginner I ask of you to point me in the right
> direction

I have a fairly extensive beginners tutorial that uses 
Python as the main language and Tcl and BASIC where 
a comparison is useful...

Try:

http://members.xoom.com/alan_gauld/tutor/tutindex.htm

Alan G.
-- 
=================================================
This post represents the views of the author 
and does not necessarily accurately represent 
the views of BT.



From andy at robanal.demon.co.uk  Sun Nov  7 11:41:19 1999
From: andy at robanal.demon.co.uk (Andy Robinson)
Date: Sun, 07 Nov 1999 16:41:19 GMT
Subject: C-Python tutorial?
References: <19991101203120.A29797@optiplex.palga.uucp> <danU3.90$AT.20506@news.shore.net>
Message-ID: <3827740a.2881032@news.demon.co.uk>

"Michael P. Reilly" <arcege at shore.net> wrote:

>Gerrit Holl <gerrit.holl at pobox.com> wrote:
>: Hello,
>
>: I have learned python as my first langauge some time ago. Is there a C
>: tutorial for python programmers, that covers C programming, espaccialy for
>: extending it with python?
>
>: I think not. Can anyone write one? I would be very grateful!
>
Someone said a long time ago that Python would be a great way to move
into learning C through writing extensions.  I feel this would be true
with the right tutorial - you get big payoffs from just a little C, as
the app framework is done for you in Python.  I don't know enough to
write this, but would sure enough review it and work with an author to
polish it up if someone had a go.

- Andy



From dave.rose at wcom.com  Fri Nov 12 10:51:53 1999
From: dave.rose at wcom.com (Dave Rose)
Date: Fri, 12 Nov 1999 15:51:53 GMT
Subject: Tkinter Label Widget and the TextVariable.
Message-ID: <tIWW3.5785$dz1.510197@pm02news>

I have assign a textvariable in the instantiation of a series of Labels.  If
I run this class by itself, the textvariables are displayed.  If I integrate
the class into a parent application, the text variables do not display even
though the variables can be printed and do hold the correct variables any
ideas?


Thanks
Dave Rose






From prestonlanders at my-deja.com  Fri Nov 26 11:56:13 1999
From: prestonlanders at my-deja.com (Preston Landers)
Date: Fri, 26 Nov 1999 16:56:13 GMT
Subject: status of remote debugging?
Message-ID: <81me3c$j51$1@nnrp1.deja.com>

Hello all,

Just wondering if there is already an implementation of remote debugging
for Python.  If not, I'm thinking it would be a neat trick.  Ideally,
what I want is to put a wrapper around my programs such that if an
unhandled exception is raised, the program will halt and listen on a
socket for remote debugging connections.  A worthwhile hack would
include simple authentication and perhaps a notification option, I
think.  Level II would include connecting to any running Python program
including threaded.

Perhaps this would require subclassing pdb to handle the client side
connection.  (IE, my local instance of pdb connects to the debugged pdb
across the network and evaluates/executes in the remote instance.)  Or,
perhaps the 'client' side of this would be a simple telnet style
interface.

I think this would be useful for debugging CGI applications, for one
thing.  It would be nice to be able to connect to the offending
application server and examine tracebacks or evaluate expressions to see
what's wrong.

Just wondering if anyone has worked on this.  I see no mention of this
in my copy of the pdb docs.

thanks,

---Preston

--
|| Preston Landers <prestonlanders at my-deja.com> ||


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



From ivanlan at callware.com  Sun Nov 14 16:24:51 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Sun, 14 Nov 1999 14:24:51 -0700
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <JlFX3.3659$s87.67930@dfiatx1-snr1.gtei.net>
Message-ID: <382F28A3.A09C3CEF@callware.com>

Hi All--

Jules Allen wrote:
> 

[snip]

> | http://www.vex.net/~x/parnassus/ is *very* nice, and it might be the
> | only solution required.  Apparently a good deal of work has been put
> | into seeding the initial database, and that's important, because it
> | makes the index immediately useful.  Yet I hadn't heard about it until
> | now; is the site still beta?  Did I miss a c.l.py.announce article
> | about it?
> 

I missed it too, and I agree; it's *very* nice.  It's a hell of a good
start.  I sure hope it's automated, and the idea of interfacing with
c.l.py.a. is a good one--a necessity, I think.

> The author was still tweaking it when he contacted me after my
> initial post. I urged him to post the URL for further inspection.
> But that's not necessary now as it's been outed!
> 
> Monday morning will be a good test of how robust it really is!
> 

Um, why Monday morning?

Boudewijn Rempt wrote:

> You don't say! I just went there, spurred on by your posting to
> enter my own efforts. Just when I had done so, I noticed that one
> of them was already there - really good work! It seems to be very,
> very nice, indeed.

Same thing happened to me.  I thought, "surely he hasn't added my stuff,
no one knows about it."  Too late, there they were.  I'm impressed.

So, who is this person?  I couldn't find any credits.  I need a link to
parnassus in my book, so I'd really like to know.

<own-up-and-fame-and-fortune-will-be-yours>-ly y'rs,
Ivan;-)
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours
----------------------------------------------



From wuhy at tcts.seed.net.tw  Fri Nov 12 18:07:37 1999
From: wuhy at tcts.seed.net.tw (Mark Wu)
Date: 12 Nov 1999 23:07:37 GMT
Subject: [HELP] TKinter could not run in Python!
References: <80i63e$7t4$1@news.seed.net.tw>
Message-ID: <80i6jp$8fl$1@news.seed.net.tw>

Dear Sir:

    Sorry, I think the statements is not clear. I think I have to explain
more detail,...
    After I install the Python, I use the prodecure to check the Tkinter..
1) import _tkinter
2) import Tkinter
3) Tkinter._test()
Then, the interpreter showed the error message that I list below...

Best Regards, Mark



Mark Wu <wuhy at tcts.seed.net.tw> wrote in message
news:80i63e$7t4$1 at news.seed.net.tw...
> Dear Sir:
>
>     This is my first time to use Python, I tried to active the Tkinter,
but
> the interpreter always tell me:
>
> error message:
> --------------------------------------------------------------------------
--
> --------------------
>
>   File "<stdin>", line 1, in ?
>   File "D:\Python\Lib\lib-tk\Tkinter.py", line 1947, in _test
>     root = Tk()
>   File "D:\Python\Lib\lib-tk\Tkinter.py", line 886, in __init__
>     self.tk = _tkinter.create(screenName, baseName, className)
> TclError: Can't find a usable init.tcl in the following directories:
>     {d:\progra~1\tcl\lib\tcl8.1} {} ./lib/tcl8.0 D:/Python/tcl8.0/library
> D:/Pyt
> hon/Tools/library
>
> d:/progra~1/tcl/lib/tcl8.1/init.tcl: version conflict for package "Tcl":
> have 8.
> 0, need 8.1
> version conflict for package "Tcl": have 8.0, need 8.1
>     while executing
> "package require -exact Tcl 8.1"
>     (file "d:/progra~1/tcl/lib/tcl8.1/init.tcl" line 19)
>     invoked from within
> "source d:/progra~1/tcl/lib/tcl8.1/init.tcl"
>     ("uplevel" body line 1)
>     invoked from within
> "uplevel #0 [list source $tclfile]"
>
>
> This probably means that Tcl wasn't installed properly.
> --------------------------------------------------------------------------
--
> --------------------
>
> My system is window NT, with TCL/TK 8.1 (already installled before install
> python)...
>
> Any suggestion and help is very appreciate :)
>
> Best Regards, Mark
>
>





From JenChu at news.vex.net  Tue Nov  2 14:55:15 1999
From: JenChu at news.vex.net (Jen Chu)
Date: 2 Nov 1999 19:55:15 GMT
Subject: Template for CGI
References: <7vn0sk$7cp$1@news.fsu.edu>
Message-ID: <7vnfj3$k1p$1@hub.org>

On Tue, 2 Nov 1999 10:47:10, Glenn Kidd <gek8365 at garnet.acns.fsu.edu> wrote:
>Is there a way to embed variables into a teplate or a replace function in
>Python?  Thank you for any help you give.

Real easy to map varables to strings in Python. I use this trick all
the time for HTML templates. You can map dictionary values into a string.
Consider the following code:

dict = {
    'text': 'html text',
    'iheight': 32,
    'HR': '<HR>',
}

textfile = """\
This isn't real %(text)s.
<IMG height="%(iheight)s">
%(HR)s
"""

print textfile % dict

----------output------------
This isn't real html text.
<IMG height="8">
<HR>









From a.eyre at optichrome.com  Tue Nov 23 12:04:12 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Tue, 23 Nov 1999 17:04:12 -0000
Subject: Python and COM: a bug?
In-Reply-To: <816010E2456BD111A48700805FBBE2EEFDF81A@ex-quebec-u1.baan.com>
Message-ID: <001f01bf35d4$c3f55f80$3acbd9c2@peridot.optichrome.com>

>> If this is vital to your project, it might be worth running the source
>> through a debugger.

> I have debugged the C++ source (our code), and I'm sure it is OK (and the
> problem never happens with lousy VBScript anyway).

Same here. Our COM object worked fine through VB.

> I have traced in the Python code, but I could not pointexactly 
> what's wrong, but I think there *is* something wrong in there.

We did get some oddities running PythonCOM through Rational's Purify: I
don't have the results any more, but I seem to remember there was internal
memory corruption somewhere. Not being a COM kind of person, I found an
easier solution by removing PythonCOM from the project entirely, and going
via an extension module instead.

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From fredrik at pythonware.com  Tue Nov 16 03:46:13 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 16 Nov 1999 09:46:13 +0100
Subject: AIFC uncompression (where's module CL?)
References: <38310AEF.93CC9BA3@speakeasy.org>
Message-ID: <010c01bf300f$0afb3060$f29b12c2@secret.pythonware.com>

Greg Whelan <gregw at speakeasy.org> wrote:
> I'm trying to use the aifc module to read a compressed AIFF-C file. The
> documentation for the library states:
> 
> "Caveat: Some operations may only work under IRIX; these will raise
> ImportError when attempting to import the cl module, which is only
> available on IRIX."
> 
> Anyone know where I can find the CL module?

it's in Lib/plat-irix[56]

if you're on a SGI, that one should be in
the path.

</F>




From achrist at easystreet.com  Wed Nov 10 23:16:48 1999
From: achrist at easystreet.com (Al Christians)
Date: Wed, 10 Nov 1999 20:16:48 -0800
Subject: ASCII delimited files
References: <942231502.1914033961@news.interbusiness.it> <382A240B.B247AD6B@python.net> <382A2801.FDE3CA09@python.net>
Message-ID: <382A4330.833C4EBD@easystreet.com>

There's a trick with files of delimited fields.  If a string field
contains a delimiter (e.g. a comma in a csv file), then the field
gets enclosed in quotes.  The handling of quotes is probably trickier,
IDK what the rules are if a field starts with or contains a single or
double quote mark.  I think I remember that some spreadsheets used tp
routinely enclose all strings in quotes in CSV files, but I'm not sure.

What I usually wind up doing to circumvent these problems with delimited
files is loading them back into a spreadsheet, typically the same one
that wrote them (I suppose that a database program would 
offer the same option), and then I write them out as tab-delimited. 
Tabs within fields are just about never in a lifetime in my work, so
far. 


Al


"Thomas A. Bryan" wrote:
> 
> "Thomas A. Bryan" wrote:
> 
> > class DelimParserField:
> >     def __init__(self, name):
> >         self.name = name
> >     def convert(self,value):
> >         return value
> >     def verify(self,value):
> >         pass
> >
> 
> One important thing that I didn't point out.  The "name" member
> is the column's name.  (Remember that I was thinking of
> delimited files for import into a database when I wrote these
> classes.)
> 
> Thus, a more sensible application might have something like
> employeeListParser = DelimFldParser((EnumField('Department',('HR','IT')),
>                                      DelimParserField('Name'),
>                                      NumericRngField('Salary',0,500000)))
> 
> If anyone uses this stuff, I'd be interested in hearing about it.
> 
> ---Tom



From mwh21 at cam.ac.uk  Sun Nov 28 15:16:02 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 28 Nov 1999 20:16:02 +0000
Subject: IDLE & Linux
References: <B466ECEC.F8E%Dan@Grassi.com>
Message-ID: <m3hfi6s77h.fsf@atrus.jesus.cam.ac.uk>

Dan Grassi <Dan at Grassi.com> writes:

> I am trying to get IDLE to run.  The docs say that it is automatically
> installed but that is just not true.  There is a /usr/bin/idle but when it
> runs it quits complaining that it can't find
> /usr/lib/python1.5/site-packages/idle/idle.py
> 
> Of course I can not find where to get the missing files.  The rpm package I
> have does not include this package.
> 
> I am running RH 6.1 and do have TCL/TK installed.
> 
> Any idea of what to do next?

Install

python-tools-1.5.2-7

you should have it knocking around with the rest of rh 6.1.

(or better, get idle from CVS or 

http://starship.python.net/crew/da/idledists/

and dump it over the one python-tools installs - it's better)

HTH,
Michael



From timmuddletin at news.vex.net  Wed Nov 24 23:56:14 1999
From: timmuddletin at news.vex.net (tim muddletin)
Date: 25 Nov 1999 04:56:14 GMT
Subject: mysql and DATE format
References: <81hhm6$ba2$1@nnrp1.deja.com>
Message-ID: <81ifhe$2igq$1@hub.org>

sp00fD at yahoo.com (sp00fD) wrote in <81hhm6$ba2$1 at nnrp1.deja.com>:
>SELECT id, subject, author, text, DATE_FORMAT(date, \"%W, %d %b %Y\"), 

Another solution is to double the % characters you don't want 
translated by python. So for (just an) example....

>>> "DATE_FORMAT(%s, \"%%W, %%d %%b %%Y\")" % 'date'
'DATE_FORMAT(date, "%W, %d %b %Y")'

Note that only the first %s was translated. All the %% are transleted
to single %. 

...




From darcy at vex.net  Thu Nov 11 17:36:48 1999
From: darcy at vex.net (D'Arcy J.M. Cain)
Date: 11 Nov 1999 22:36:48 GMT
Subject: WEB DATABASE PROGRAMMER POSITION AVAILABLE
References: <MuFW3.102$yd1.9175@typhoon-la.pbi.net> <80fdts$fml$1@nntp1.atl.mindspring.net>
Message-ID: <80fge0$rkv$1@hub.org>

Aahz Maruch <aahz at netcom.com> wrote:
> In article <MuFW3.102$yd1.9175 at typhoon-la.pbi.net>,
> David Winsen <david at hotjobs2000.com> wrote:
>>
>>WEB DATABASE PROGRAMMER POSITION AVAILABLE
>>URGENT MESSAGE!

> Who wants to net.cop this one, or should I do it?

I thought we might all get together and send back a community resume with
all our experience and training.  :-)

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



From fredrik at pythonware.com  Thu Nov  4 03:24:43 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 4 Nov 1999 09:24:43 +0100
Subject: Comparison between Python and "Ruby"
References: <1270552684-31022246@hypernet.com> <87n1svugfh.fsf@ev.netlab.co.jp>
Message-ID: <012101bf269e$0ce118b0$f29b12c2@secret.pythonware.com>

Yukihiro Matsumoto <matz at netlab.co.jp> wrote:
> But TMTOWTDI (in other words: the more, the merrier) is the slogan to
> emphasize maximum variability in ALL AREAS, so that I felt weird to
> see Pythoneers talking TMTOWTDI or even TMTOWTHMTOWTDI in positive
> attitude.  Reducing variablility in ANY AREA contradict against
> TMTOWTDI, I think.

well, I can think of many areas where Perl
doesn't give you maximum variability when
compared to other languages (of all kinds).

(in other words, the TMTOWTDI argument is
nonsense, just as I've always suspected).

</F>




From tim_one at email.msn.com  Tue Nov  2 00:05:30 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Tue, 2 Nov 1999 00:05:30 -0500
Subject: __str__ vs. __repr__
In-Reply-To: <1270634443-26149069@hypernet.com>
Message-ID: <000701bf24ef$e2d29b20$0f2d153f@tim>

[Randall Hopper gives an example]
>    class c:
>        def __str__(self):  return 'foo'
>        def __repr__(self): return 'bar'
>
>    >>> str(c())
>    'foo'
>
>    >>> str([c(), c(), c()])
>    '[bar, bar, bar]'

[which Gordon McMillan explains, adding]
> Sick enough, but not convoluted enough, to be a Tim-ism...

It's too convoluted for my tastes!  The real illness is that lists (and
dicts, and tuples) don't pass str-vs-repr'ness *down*.  That is, even though
the example explicitly asks for str of a list, the list object asks for the
repr of its elements.  That's convoluted:  if (as happens to be true) str is
meant to give a friendly string, why do the builtin container types ask
their containees to produce unfriendly strings regardless?  For that matter,
why is repr implied at an interactive prompt given a raw expression?  Seems
that friendly strings would be more appropriate there.  And why is there a
shorthand (`x`) for repr(x) but not for str(x)?

These are Great Mysteries debated even among the Ancient Ones.  If the
scheme didn't already exist, I don't think we'd have much luck arguing for
its adoption <wink>.

3/4ths-of-one-idea-plus-1/4th-of-another-adds-up-to-2/3rds-ly y'rs  - tim





From JamesL at Lugoj.Com  Thu Nov 25 17:29:18 1999
From: JamesL at Lugoj.Com (James Logajan)
Date: Thu, 25 Nov 1999 14:29:18 -0800
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <867lj85g4c.fsf@g.local> <383CC835.B585EBDD@Lugoj.Com> <00bd01bf3723$36d699c0$f29b12c2@secret.pythonware.com>
Message-ID: <383DB83E.F985AED@Lugoj.Com>

Fredrik Lundh wrote:
> 
> James Logajan <JamesL at Lugoj.Com> wrote:
> > Sorry, but I think the price paid for "true" GC is too high for the small
> > number of programs that benefit from this. Perhaps one should learn to clean
> > up after oneself when cycles are expected.
> 
> Hmm.  I've recently dissected two large (and very complicated)
> Python projects, and one of the major cost contributors when
> going from "working code" to "production quality code" was the
> time needed to hunt down a number of totally unexpected re-
> source leaks.  GC would have helped in most of those cases.
> 
> (and in contrast what people claim on this list, more
> syntactic sugar or static typing would have saved al-
> most nothing...)
> 
> I plan to write a longer piece on this at a later time.

But you were able to track those resource leaks down. And correct them
because the language was designed with enough flexibility to aid in a
solution, not designed with a perhaps inappropriate and unchangeable
a-priori solution. Consider:

I've worked on two projects where deterministic memory management was an
implied requirement. One was a field-programmable embedded device for use in
remote network monitoring (actually, see
http://www.badger.com/html/main/products/oremote.html although this doesn't
give a lot of information on the software). Java was considered and dropped
for several reasons. Its non-deterministic GC didn't help its case at all.
If Python had had non-deterministic GC behavior, we might not have chosen it
either. Since it is programmable, it is possible for technicians to write
Python code with memory leaks. But memory leaks in application scripts can
be found and fixed and memory recovered at known points in processing.
However, a GC algorithm built into a language that is inappropriate for a
given application allows for no fixing...ever.

The other project is a telecommunications device for a startup firm where I
work now. Java and C++ were both considered as possible tools for the bulk
of development. Most developers on the team were in favor of Java because
newer compilers and VMs were closing the speed gap with C++ and Java's
memory management would have sped up development. Alas, after discussing our
requirements with Sun Java experts we came away with two important
conclusions:

1) Amazingly, Sun couldn't guarantee that their Java VM didn't have leaks!
2) They strongly recommended against using Java in any mission-critical
applications where non-deterministic GC behavior would be detrimental to the
application users (i.e. sudden execution pauses while the VM did a
mark-and-sweep collection).

We do use Java some, but only outside the core application which is written
in C++.

I suppose any solution for Python that allows some flexibility in the matter
would be nice. Perhaps keep the reference counting and add interpreter flags
to allow a choice of other GC schemes on top of that to clean up cycles.

Just my 2 cents....



From edwardam at home.com  Mon Nov 29 23:57:00 1999
From: edwardam at home.com (Edward Muller)
Date: Tue, 30 Nov 1999 04:57:00 GMT
Subject: Exposing COM via XML-RPC or Something Else
Message-ID: <38435919.D88B4EC0@home.com>

I'm playing around with a way to expose Windows COM object via
XML-RPC (or something else).

Let me explain:

My first thought was to write an XML-RPC API to expose the MAPI COM
object and it's methods. This way I (or others) could write a
MAPI client on other platforms (Linux, Web, Java, etc) and these would
talk to my XML-RPC-MAPI server (Lets ignore MAPI Profiles for a second,
jsut for arguments sake). As I thought about the project a little more I
figured that Microsoft already has an object heirarchy and API built,
why should I recreate the wheel, adding another layer of complexity (and
another API) ontop. So I started to look deeper into the XML-RPC
protocol and Python xmlrpclib, trying to figure out a way to expose the
COM objects via XML-RPC. I realized (I should have already know this,
but I was being a DUH) that XML-RPC doesn't have any idea about Objects,
just simple data types (int, short, long, string, etc, etc).

So my current question, can anyone think of an existing technology or
library that will allow me to expose these COM objects in a way that
could be used by other platforms/languages. Corba (FNORB), ILU, PYRO?

Or am I really just going a little crazy?

Should I just stick to my original idea? Writing an XML-RPC API exposing
the MAPI COM objects. This was someone could write a client using the
XML-RPC API to access MS Exchange Servers from Linux, or a Web Server or
a Java Client, et. al. OR should I shot for the stars and try to write
an abtraction layer allowing us to expose MS COM objects with something
like Corba, ILU or PYRO? This (if it worked) would allow us all to use
COM objects from other platforms that support whatever library I've used
to abstract/expose the COM object(s). This all assumes someone is
running a COM server on the machine the client would connect to...


...EAM...




From singer at itl.atr.co.jp  Tue Nov 30 09:08:59 1999
From: singer at itl.atr.co.jp (Harald Singer)
Date: Tue, 30 Nov 1999 23:08:59 +0900
Subject: COM and Powerpoint control
Message-ID: <3843DA7B.60CE99B2@itl.atr.co.jp>

Hi,

I am looking for an example of an "outside controlled" PowerPoint
presentation using COM. My final goal is to time-align the slides
presentation with an audio file, because the PowerPoint internal
slide-show timing is massively broken.

The following code snippet worked promisingly well on Win98:

 from win32com.client import constants, Dispatch

 x = Dispatch('Powerpoint.Application.8')
 x.Presentations.Open(FileName='c:\temp\COMTutorial.ppt')
 x.ActiveWindow.ViewType = constants.ppViewNotesPage

Now, I tried to record a macro to see what commands might be used in
slideshow mode, but looking at the generated VB code, I can't see any
commands that advance the slideshow.

-- VB snippet start ---
Sub Macro1()
'
 ppShowAll
        .AdvanceMode = ppSlideShowUseSlideTimings
        .PointerColor.SchemeColor = ppForeground
        .Run
    End With
    SlideShowWindows(Index:=1).View.Exit
End Sub'
-- VB snippet end ---

Any help or code fragments are appreciated. BTW, I did a search
for this but nothing relevant came up.
-- 
Harald Singer	ATR-ITL		singer at itl.atr.co.jp	++81-774-95-1389




From j.nijtmans at chello.nl  Thu Nov  4 14:49:25 1999
From: j.nijtmans at chello.nl (Jan Nijtmans)
Date: Thu, 04 Nov 1999 20:49:25 +0100
Subject: embedding stubbed tcl
References: <bLd7vSA1euH4EwDS@jessikat.demon.co.uk>
	 <8ZCT3.18814$1v5.202282@newscontent-01.sprint.ca> <VVo0SGAJExH4EwyM@jessikat.demon.co.uk>
Message-ID: <3821E345.FC6C372F@chello.nl>

Robin Becker wrote:

> I thought stubs were for extensions. The tkinter case is slightly the
> wrong way round as it starts tcl up.

Stubs can very well be used for applications as well. For example,
I already used it to Stubify the Tcl plugin, so I don't need to
recompile it when Tcl8.3 comes out ;-).  The only problem is that
Tcl doesn't yet contain a portable way to use dlopen()/LoadLibrary()
or whatever this function might be called.

-- 
Jan Nijtmans, CMG Arnhem B.V.
email: j.nijtmans at chello.nl (private)
       jan.nijtmans at cmg.nl  (work)
url:   http://purl.oclc.org/net/nijtmans/



From aahz at netcom.com  Wed Nov 24 11:08:17 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 24 Nov 1999 16:08:17 GMT
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <81g8po$reo$1@mach.vub.ac.be>
Message-ID: <81h2hh$85c$1@nntp9.atl.mindspring.net>

In article <81g8po$reo$1 at mach.vub.ac.be>,
Thomas Hamelryck <thamelry at vub.ac.be> wrote:
>
>Block delimitation by indenting. This makes python look like a porsche
>with a caravan attached :-). As far as I know, most python users even
>seem to like this for some reason.

I agree with you and I'm also one of the people who likes it.  Here's
why: it makes Python code *readable*.  In fact, I wish that Guido would
disallow

  if foo : bar()

and force it to always be

  if foo :
    bar()
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)

Plus ca change, plus c'est la meme chose.



From pinard at iro.umontreal.ca  Mon Nov  1 18:24:20 1999
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 01 Nov 1999 18:24:20 -0500
Subject: Improved struct module
In-Reply-To: "Robin Boerdijk"'s message of "Thu, 14 Oct 1999 20:47:43 +0200"
References: <000001bf15fc$3397ee00$ba2d153f@tim> <0E16861EE7BCD111BE9400805FE6841F0BEB2650@c1s5x001.cor.srvfarm.origin-it.com>
Message-ID: <oqso2phk2j.fsf@titan.progiciels-bpi.ca>

"Robin Boerdijk" <robin.boerdijk at nl.origin-it.com> writes:

> I must admit that I am also one of those lazy bastards that uses his
> newsreader's "find any messages by Tim Peters"-function to locate
> interesting threads.

So do I, but even more lazily!  Gnus bolds Tim's messages automatically
for me, when preparing the threaded summary for the Python mailgroup. :-)

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





From 71351.356 at compuserve.com  Wed Nov 17 01:00:18 1999
From: 71351.356 at compuserve.com (T. C. Mits)
Date: Wed, 17 Nov 1999 01:00:18 -0500
Subject: Need help dealing with spaces in NT paths.
References: <80pr76$jq5$1@nnrp1.deja.com> <11f733ec.96958819@usw-ex0101-007.remarq.com> <uogcu93dt.fsf@cs.uu.nl> <80sj4q$l7t$1@nnrp1.deja.com>
Message-ID: <80tgi8$hbn$1@ssauraaa-i-1.production.compuserve.com>

A simple example of the difficulty is:

>>> os.system(r'"dir Program Files"')
Bad command or file name
0
>>>


This is run at the root of C: for simplicity.


I also get:

os.popen(r'"dir Program Files"', "r")
Bad command or file name
<open file '"dir program files"', mode 'r'  at 795930>

What does that mean?
----------------------------------------------------------------------------
---------------------------------
Now this is weird.   I created a test.bat file that just has dir %1 in it.

>>> import os
>>> os.execv("test.bat",["program files"])

C:\>dir files

 Volume in drive C is CDRIVE
 Volume Serial Number is 2E50-0FE9
 Directory of C:\

File not found
                      335,601,664 bytes free

C:\>

Looks like it blew the interpreter too.




os.system("dir Program Files")
<maytagrepairman at yahoo.com> wrote in message
news:80sj4q$l7t$1 at nnrp1.deja.com...
> Ok.  This works so far, but to complicate things further I need also to
> pass an argument to the command.  Both the command and the argument are
> full path names with spaces in them.  If I type something like this at
> the DOS prompt:
>
> "my_full_path_cmd_with_spaces" "my_full_path_arg_with_spaces"
>
> it works.
>
> But if I try to do an equivalent such as this in Python
>
> # this next statement is to see what is passed to os.system()
>     print ('\"%s\" \"%s\"' % (cmd,arg))
>
> # this next statement is the call to os.system()
>     os.system('\"%s\" \"%s\"' % (cmd,arg))
>
> I get
>
> The name specified is not recognized as an
> internal or extenal command, operable program or batch file.
>
> Other variations I've tried give the same results.  Any thoughts on
> this one?
>
>
> In article <uogcu93dt.fsf at cs.uu.nl>,
>   piet at cs.uu.nl wrote:
> > >>>>> wheineman <wheinemanNOwhSPAM at uconect.net.invalid> (W) writes:
> >
> > W> Put quotes around the complete path. For example:
> > W> myExec = "C:\\Program Files\\Trash Can\\compactor.exe"
> >
> > You probably mean:
> > myExec = "\"C:\\Program Files\\Trash Can\\compactor.exe\""
> >
> > or easier:
> > myExec = r'"C:\Program Files\Trash Can\compactor.exe"'
> > --
> > Piet van Oostrum <piet at cs.uu.nl>
> > URL: http://www.cs.uu.nl/~piet [PGP]
> > Private email: Piet.van.Oostrum at gironet.nl
> >
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.





From robin at alldunn.com  Thu Nov 11 21:52:34 1999
From: robin at alldunn.com (Robin Dunn)
Date: Thu, 11 Nov 1999 18:52:34 -0800
Subject: Beginning programming with Python
References: <sng0ychgwt.fsf@motorola.com> <14379.17276.267502.673474@dolphin.mojam.com>
Message-ID: <tiLW3.437$0O.9764916@news.randori.com>

>
>     Aaron> Secondly, if I choose to use Python for this, I would like to
>     Aaron> supply each student with a reference book on Python.
>
> From what I've seen on the list, Learning Python sounds like a more
> appropriate book for first-time programmers, though I've personally not
seen
> it.

I am currently teaching a Python class to a group of programmers for whom
OOP is a very foriegn concept, (so most of them are ALMOST in the same boat
as first-time programmers...)  I specified Learning Python as the text book
and have been teaching from it, assigning the exercises from it, etc.  I
gave the first test yesterday and everybody did quite well.  i have to give
the book a lot of credit for that.  Most of the concepts covered so far are
new to these guys, but with the exception of a few questions here and there,
they are getting them down after reading and a summary type lecture.


--
Robin Dunn
Software Craftsman
robin at AllDunn.com
http://AllDunn.com/robin/
http://AllDunn.com/wxPython/  Check it out!






From d.huesken at gmx.de  Thu Nov 11 14:53:21 1999
From: d.huesken at gmx.de (Dirk Huesken)
Date: Thu, 11 Nov 1999 20:53:21 +0100
Subject: Image buttons with Tkinter
References: <38270596.A391995C@parlant.com>
Message-ID: <hr6f08.093.ln@amoeba>

Thomas Lane <tlane at parlant.com> wrote:
>
>Also, is there a place on the net where I can find lots of sample
>Python/Tkinter code? I've read Fredrik Lundh's intro, but I haven't been
>able to find any other code or reference material for Tkinter.

What about Python-1.5.2/Demo/* from the Python source distribution
(especially Python-1.5.2/Demo/tkinter)?

-- 
Dirk H?sken
:wq



From zakons at eccelerate.com  Wed Nov 10 16:16:52 1999
From: zakons at eccelerate.com (zakons at eccelerate.com)
Date: Wed, 10 Nov 1999 21:16:52 GMT
Subject: what is a function object?
References: <809p0a$8nf$1@nnrp1.deja.com> <m31z9znxy4.fsf@atrus.jesus.cam.ac.uk> <80bv3n$s9b$1@nnrp1.deja.com> <m3iu3amkkz.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <80cnbv$g30$1@nnrp1.deja.com>

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

> Then "re.sub('{{[^}]*}}',Replacer(n).replace,line)" does what I want
> (NB: I could have called the `replace' method `__call__' and then
> written `Replacer(n).replace' as `Replacer(n)'; that I didn't is a
> personal bias of mine and makes no difference to the point).
>

Cool example Michael. I definitely prefer the use of `__call__' in this
example because it would encapsulate the entire behavior a lot better. I
like the idea of instantiating an object with behavior determined by the
parameters passed to the constructor, and having that object implement
the function's logic through the saved state.

Thanks,
Stuart



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



From mhammond at skippinet.com.au  Thu Nov  4 05:00:10 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 4 Nov 1999 21:00:10 +1100
Subject: Embedding and threads
In-Reply-To: <00b501bf2699$567dccc0$4500a8c0@thomasnotebook>
Message-ID: <001b01bf26ab$61c951f0$0501a8c0@bobcat>

This is a huge source of confusion, and I cant remember the rules any
more.

For working code, see the "CEnterLeavePython" class defined in the
pywintypes module (I can send you the source if you dont have them
handy).

For a long and protracted discussion about this and much more, see the
thread-sig archives from around 12 months ago - this was discussed in
so much detail it made my nose bleed :-)

Mark.

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Thomas Heller
> Sent: Thursday, 4 November 1999 6:51
> To: python-list
> Subject: Embedding and threads
>
>
> While embedding python in a DLL I'm writing (we are talking
> Windows here),
> I encountered the following confusion:
>
> Threads are created in the application from C. From reading the docs
I
> understood that I must 'boot the thread' into existance by
> creating a thread
> state data structure (with PyThreadState_New(), passing the
> PyInterpreterState *interp pointer which I saved from the
> main thread),
> then acquire the interpreter lock, and finally store the thread
state
> pointer before using the Python/C API from this thread.
>
> I do this in the DllMain() function in the DLL_THREAD_ATTACH branch.
>
> This all works very well, but what happens if I also create threads
> from python? These threads will be initialized by python, so
> I assume the thread state data structure is already created.
>
> How do I ask python if the thread state already exists for a
> given thread?
>
> Thomas Heller
> ION-TOF GmbH
>
>
> --
> http://www.python.org/mailman/listinfo/python-list
>




From hinsen at cnrs-orleans.fr  Tue Nov  2 04:57:07 1999
From: hinsen at cnrs-orleans.fr (Konrad Hinsen)
Date: 02 Nov 1999 10:57:07 +0100
Subject: ANN: small vector operation module
References: <3815FD36.F43DFECF@mail.medscape.com> <7v5fuu$1692@news.rhrz.uni-bonn.de> <Pine.SOL.4.05.9910261806090.4705-100000@venus.radsci.uci.edu>
Message-ID: <m3yachqkr0.fsf@chinon.cnrs-orleans.fr>

Fuming Wang <fuming at venus.radsci.uci.edu> writes:

> I wrote a small module that does vector operations. It's quite useful for
> processing lab data. Comments are welcome.

Looks OK, but why don't you just use Numerical Python? It offers much
more and is more efficient as well.
-- 
-------------------------------------------------------------------------------
Konrad Hinsen                            | E-Mail: hinsen at cnrs-orleans.fr
Centre de Biophysique Moleculaire (CNRS) | Tel.: +33-2.38.25.55.69
Rue Charles Sadron                       | Fax:  +33-2.38.63.15.17
45071 Orleans Cedex 2                    | Deutsch/Esperanto/English/
France                                   | Nederlands/Francais
-------------------------------------------------------------------------------



From brian at liaone.ne.mediaone.net  Sun Nov 14 16:42:19 1999
From: brian at liaone.ne.mediaone.net (brian at liaone.ne.mediaone.net)
Date: Sun, 14 Nov 1999 21:42:19 GMT
Subject: Other ways to disable Wm in Tkinter?
Message-ID: <80nabr$q62$1@nnrp1.deja.com>

Hello,
I have a question on how to dynamically activate/deactivate the window
manager. I'm using a button activated listbox to enter items into an
application. The listbox is on a separate managerless Toplevel to give
it a placard appearance.  When the listbox pops up, it has a local grab.
Unfortanately, one can still drag around the main window underneath
and thus inadvertantly hide the listbox.  I have managed to get around
this problem by using withdraw/overrideredirect/deiconify on the
main window when the listbox appears and again when the listbox
disappears (to restart the window manager).  This is not a good fix
since it results in irritating flashes as the main window is
clobbered and redrawn.

Thus, my question - can one disable the window manager (or at least
disable window movement) short of withdraw/overrideredirect/deiconify?
It goes without saying that using grab_set_global is a bit too brutal.
Thanks,
Brian

------------------here is some example code-------------------------

#!/usr/bin/env python

import sys
from Tkinter import *

controller_list = ['a','b','c','d','e','f','g','h','i','j','k','l','m',
                   'n','o','p','q','r','s','t','u','v','w','x','y','z']

class Right(Frame):
    def __init__(self,master,big):
        self.big = big
        Frame.__init__(self)
        self.right = Frame(master)
        self.right.pack(side=RIGHT)
        for i in range(10):
          for j in range(10):
            self.b = Button(self.right,bg='#44d950',width=8,
                                 height=1,text='button ' + `i` + `j`)
            self.b['command']=lambda cblb=self.callbackCreateListbox,\
                                 r = self.b : cblb(r)
            self.b.grid(row=i,column=j)
    def callbackDestroyListbox(self):
        self.temp.destroy()
    def handleListChoice(self,event):
        label = self.temp.list.get(ACTIVE)
        self.controller_entry['text']=label
        self.temp.destroy() #now we destroy listbox
        self.big.turnOnWm() #now to re-enable the Window Manager
    def callbackCreateListbox(self,controller_entry):
        self.controller_entry = controller_entry
        x0 = controller_entry.winfo_rootx()
        y0 = controller_entry.winfo_rooty() +       \
                           controller_entry.winfo_height()
        self.temp = Toplevel()
        self.temp.geometry("+%d+%d" % (x0, y0))
        self.temp.overrideredirect(1)
        self.temp.scroll = Scrollbar(self.temp)
        self.temp.list = Listbox(self.temp,height=15,width=10,
                                yscrollcommand=self.temp.scroll.set)
        self.temp.list['selectmode']=SINGLE
        self.temp.list.grid(column=0)
        pos = 0
        for controller in controller_list:
            self.temp.list.insert(pos,controller)
            pos = pos+1
        self.temp.scroll.config(command=self.temp.list.yview)
        self.temp.scroll.grid(row=0,column=1,sticky=N+S)
        self.temp.button = Button(self.temp,text='Quit',command=
                                  self.callbackDestroyListbox)
        self.temp.list.bind('<Double-Button-1>',self.handleListChoice)
        self.temp.grab_set() #Almost does the trick-unfortunately Window
                             #manager is unaffected.  Thus, must
                             #withdraw, overrideredirect and deiconify.
                             #Note: using grab_set_global works but is
                             #too brutal.
        self.big.turnOffWm(self.temp.list) #need to give turnOffWm() a
                                           #ref back to this listbox
                                           #for stacking purposes.
class Big(Frame):
    def __init__(self):
        self.placardCnt='even' #even/odd cnt of listboxes
        root = Tk()
        root.withdraw()
        Frame.__init__(self)
        self.top = Toplevel(width=100,height=
                           500,bg='red') #create master Frame
        self.top.protocol("WM_DELETE_WINDOW",root.quit)
        self.right = Right(self.top,self)
    def turnOnWm(self):
        heightOfWm     = -25
        widthOfWmFrame = -5
        x0 = self.top.winfo_x() + widthOfWmFrame
        y0 = self.top.winfo_y() + heightOfWm
        self.top.withdraw()
        self.top.overrideredirect(0)
        self.top.geometry("+%d+%d" % (x0,y0))
        self.top.deiconify()
    def turnOffWm(self,listboxToplevel): # need a handle on listbox to
                                         # place main window(self.top)
                                         # below listbox.
        if self.placardCnt=='even':
            heightOfWm    = 1 #Should =0,but then toplevel
                              #reappears too high up.
            self.placardCnt='odd'  #'increment' placard count
        else:
            heightOfWm    = -1 #A kludge to keep toplevel from slowly
                               #creeping downward.
            self.placardCnt='even' #'increment' placard count
        widthOfWmFrame = 0
        x0 = self.top.winfo_x() + widthOfWmFrame
        y0 = self.top.winfo_y() + heightOfWm
        self.top.withdraw()
        self.top.overrideredirect(1)
        self.top.update_idletasks()
        self.top.geometry("+%d+%d" % (x0,y0))
        self.top.update_idletasks()
        self.top.lower(belowThis=listboxToplevel) #keeps listbox from
                                                  #being obscured.
        self.top.deiconify()

if __name__ == '__main__':
    b = Big()
    b.mainloop()



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



From moonseeker at my-deja.com  Sun Nov 28 08:03:25 1999
From: moonseeker at my-deja.com (moonseeker at my-deja.com)
Date: Sun, 28 Nov 1999 13:03:25 GMT
Subject: AIX Py compiling errors?
Message-ID: <81r96t$meq$1@nnrp1.deja.com>

Hi!

Anyone using Python on AIX here? I tried to compile Python on my AIX 4.1.5
box using the IBM compiler but I get an error (or warning):

  cc_r	-O -qmaxmem=4000 -I./../Include -I.. -DHAVE_CONFIG_H -c
./socketmodule.c "./socketmodule.c", line 1127.22: 1506-280 (W) Function
argument assignment between types "struct sockaddr*" and "unsigned char*" is
not allowed.

Anyone a suggestion?

Mike


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



From tlane at parlant.com  Tue Nov  9 10:56:35 1999
From: tlane at parlant.com (Thomas Lane)
Date: Tue, 09 Nov 1999 08:56:35 -0700
Subject: Grail Brower question
References: <808m9h$nnn$1@holly.prod.itd.earthlink.net>
Message-ID: <38284433.5AF3B00E@parlant.com>

Have you installed Python yet?

Spaceram wrote:
> 
> For anyone who knows what the grail browser is I have a question.
> How do I use the brower? I download it and when i doubleclick it it opens a
> open with box which means that windows doesn't know what to do with the
> file.
> 
> Please email all answers to spaceram00 at earthlink.net

-- 
Thomas Lane
Provo, UT
tlane at parlant.com



From mhammond at skippinet.com.au  Sun Nov  7 17:21:09 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sun, 07 Nov 1999 22:21:09 GMT
Subject: win32com and IIS?
References: <802s6a$mj5$1@web1.cup.hp.com> <Q_7V3.5012$pf4.18247@news-server.bigpond.net.au> <803e2a$dul$1@web1.cup.hp.com>
Message-ID: <pXmV3.15856$pf4.21246@news-server.bigpond.net.au>

As you guess, this is almost certainly a permssions problem.  IIS may not
have permissions to create windows on its workstation.  The error message
certainly implies this is indeed the case.  The only thing I can suggest is
searching the MS knowledge base, and see what they have to say - I am fairly
sure that _all_ languages that can run under IIS will have the same issue.

As a simple test, you could use the Windows Scripting Host from CGI, and do
the same thing with some VBScript code.  I feel confident it will have the
exact same problem...

Mark.

Arad Rostampour wrote in message <803e2a$dul$1 at web1.cup.hp.com>...
>Thanks for the quick reply...Here is a simpler example with debugging flags
>on:
>
>import win32com.client
>ieweb = win32com.client.Dispatch("InternetExplorer.Application")
>ieweb.Navigate("http://www.hp.com")
>print ieweb.Document.Body.innerText
>
>Attribute Document not in cache
>Getting property Id 0xcb from OLE object
>Traceback (innermost last):
> ...
>pywintypes.com_error: (-2147023169, 'The remote procedure call failed and
>did not execute.', None, None)
>
>Again, it works from the python interactive interpreter but does not work
in
>a cgi script under IIS.  It doesn't like getting ieweb.Busy either...Also,
>are there any issues with running the COM interface not specifically as
>myself or a specific user, but rather IUSR_machinename?  Is there any
>special set-up besides the Script Map key that ties a .py file to python?
>






From bernhard at alpha1.csd.uwm.edu  Tue Nov 30 13:06:42 1999
From: bernhard at alpha1.csd.uwm.edu (Bernhard Reiter)
Date: 30 Nov 1999 18:06:42 GMT
Subject: COMobject problem
References: <3843e3f3.18656326@122.0.0.250>
Message-ID: <slrn8484hh.cn.bernhard@alpha1.csd.uwm.edu>

On Tue, 30 Nov 1999 14:51:54 GMT, volumeregeling <volumeregeling at hotmail.com> wrote:
>i have a problem regarding COM objects, i invoke the object through
>the following statement :
>
>Editor= win32com.client.Dispatch("Comeditor.Classeditor")
>
>Now i have the problem that when i close the main program , the
>COMobject will still be alive, i want to close all the comobjects
>opened from within python to be closed when the main program closes.
>
>I have no idea on how to do that, maybe you do , then i will be very
>gratefull.

You could delete the Editor variable and hope that it cleans itself up.
Usually there is a special method to be called to close the application.

If you are using the win32com extensions, you can try to browse the COM library
(applying the COM makepy.py script on the type library helps for that).

Like with SPSS here there is a Quit() method in the main application object.

So Editor.Quit() with Quit replacing the Method your are looking for, should 
work.

Just my 0.02 Euro.
	Bernhard
-- 
Research Assistant, Geog Dept UM-Milwaukee, USA.  (www.uwm.edu/~bernhard)
Free Software Projects and Consulting 		         (intevation.net)  
Association for a Free Informational Infrastructure            (ffii.org)



From Bjoern.Giesler at stud.uni-karlsruhe.de  Tue Nov 23 13:36:13 1999
From: Bjoern.Giesler at stud.uni-karlsruhe.de (Bjoern.Giesler at stud.uni-karlsruhe.de)
Date: Tue, 23 Nov 1999 19:36:13 +0100 (CET)
Subject: Embedding: Again
Message-ID: <E11qKnY-0005dV-00@rzstud1.rz.uni-karlsruhe.de>

Hi,

OK, nevermind, I think I've got it: In CPluginPyRegister(), I use
(theController is the plugin instance):

  /* Here we go: Obtain module name... */
  moduleName = PyObject_GetAttrString(theController, "__module__");
  if(!moduleName) return NULL;
  /* ..., module dictionary... */
  modulesDict = PyImport_GetModuleDict();
  if(!modulesDict) return NULL;
  /* ...and module object */
  theModule = PyObject_GetItem(modulesDict, moduleName);
  if(!theModule) return NULL;

Now my only problem is, which of these pass on ownership? Meaning, what do I
have to do to truly unload a module? Is Py_XDECREF(theModule) sufficient, or
must I do a PyObject_DelItem(modulesDict, moduleName) as well?

Thanks in advance,
		Bjoern
-- 
--------------------------------/\--One OS to rule them all---Windows NT-----
Bjoern Giesler                 /  \  One OS to find them           
<un4e at rz.uni-karlsruhe.de>    / <> \  One OS to bring them all    
-----------------------------/______\--And in the Darkness bind them---------
    Thank you for your time, worship the Antichrist, and have a nice day.



From pinard at iro.umontreal.ca  Mon Nov  1 19:48:33 1999
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 01 Nov 1999 19:48:33 -0500
Subject: Default indentation of 4 columns
In-Reply-To: Gareth McCaughan's message of "24 Oct 1999 20:43:12 +0100"
References: <oq904sc08y.fsf@titan.progiciels-bpi.ca> <86emeko88f.fsf@g.local>
Message-ID: <oqeme9hg66.fsf@titan.progiciels-bpi.ca>

Gareth McCaughan <Gareth.McCaughan at pobox.com> ?crit:

> Fran?ois Pinard wrote:

> > The only thing that really rebukes me is the interdiction to use
> > capitals in comments to mark non-terminals (or variables), as it does
> > not get accompanied by a better suggestion, or at least, none yielding
> > as much clarity.

> You mean saying things like

>     # Return the sum of AARDVARK and ARMADILLO.
>     def f(aardvark, armadillo):
>         return aardvark+armadillo

> I use a convention derived from Knuth's literate programming
> systems, and write

>     # Return the sum of |aardvark| and |armadillo|.

It is a good idea.  Yet, maybe because I've been so exposed to GNU software,
I'm quite used to the capitals convention, and find it very legible.
I would have tendency to stick to something in (relatively) wide use, than
to run away from the dilemma and hide into something less well known :-).

> ? (Incidentally, this example is another one that shows an
> annoying "accidental alignment" as a result of the four-character
> rule.

Usually, the `def' is outstanding enough already, preceded by a blank line,
being left aligned, and especially highlighted in some good editors (:-),
that I would not think the above really creates a real practical problem.

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





From MSteed at altiris.com  Mon Nov  8 11:23:08 1999
From: MSteed at altiris.com (Mike Steed)
Date: Mon, 8 Nov 1999 09:23:08 -0700 
Subject: Python and generating images for web
Message-ID: <65118AEEFF5AD3118E8300508B124877073C68@webmail.altiris.com>

> From: Alexander Stepanov [mailto:a_stepanov at yahoo.com]
> Sent: Saturday, November 06, 1999 8:53 PM
> To: python-list at python.org
> Subject: Python and generating images for web
>
> [snip]
> I will appreciate if you point me out accessible chart/plot  
> modules and
> libraries as well as documentation and
> working sites.    Any additional information is strongly welcome.

I wrote a Python interface for Bruce Verderaime's GDChart library, available
here:

http://members.xoom.com/msteed/software/gdchart.html

In its current form, GDChart generates GIF images, but it can be changed to
generate PNG if you prefer.  (This would require updating the gd library on
which GDChart is built, and perhaps some minor tweaks to GDChart itself.)

The archive includes a precompiled Win32 binary.

Mike.



From info at objectsbydesign.com  Wed Nov 10 22:31:35 1999
From: info at objectsbydesign.com (Stuart Zakon)
Date: Wed, 10 Nov 1999 22:31:35 -0500
Subject: UML tools: Tcl or Python?
Message-ID: <382A3896.7FD4F3B0@objectsbydesign.com>

Interesting developments for scripting in UML tools:

First ObjectDomain switched its scripting language from Tcl to JPython.
Next Together/J 3.0 switched its scripting language from JPython to Tcl.

Now that's what I call product differentiation!!

Stuart Zakon
Objects by Design
http://www.objectsbydesign.com





From quinn at groat.ugcs.caltech.edu  Fri Nov 26 21:43:58 1999
From: quinn at groat.ugcs.caltech.edu (Quinn Dunkan)
Date: 27 Nov 1999 02:43:58 GMT
Subject: [].extend only accepts list?
Message-ID: <slrn83uhd2.6lc.quinn@groat.ugcs.caltech.edu>

Is there any reason why the extend method of lists shouldn't accept any
sequence?



From eric at net.tamu.edu  Wed Nov  3 16:16:54 1999
From: eric at net.tamu.edu (Eric Nelson)
Date: Wed, 3 Nov 1999 15:16:54 -0600
Subject: 64 bit ints for SNMP counter64 variables
References: <7vq8eg$4k$1@news.tamu.edu>
Message-ID: <7vq8o4$18j$1@news.tamu.edu>

oops... posted too quick... found long() just after i hit send... sorry

eric

Eric Nelson <eric at net.tamu.edu> wrote in message
news:7vq8eg$4k$1 at news.tamu.edu...
> Is there not a 'long long' data type in python? I've got 64 bit counters
> coming from snmp variables that I need to work with....
>
> How about a variable precision type module??
>
> Eric
> eric at net.tamu.edu
>
>





From steve at acer.spvi.com  Sun Nov  7 22:33:01 1999
From: steve at acer.spvi.com (Steve Spicklemire)
Date: Sun, 7 Nov 1999 22:33:01 -0500 (EST)
Subject: Python for children.....
In-Reply-To: <199911080328.WAA24200@acer.spvi.com> (message from Steve
	Spicklemire on Sun, 7 Nov 1999 22:28:21 -0500 (EST))
References: <9911051740130D.02530@william.jinsa.org> <7GOU3.272$Ae3.194028@news-dal.corridex.com> <801p16$10s$1@vvs.superst.iae.nl> <199911080328.WAA24200@acer.spvi.com>
Message-ID: <199911080333.WAA24231@acer.spvi.com>

Forgot to mention:

PyBots: http://opensource.spvi.com

Also.. there is a windows version, just like the Mac, but there is no
rotation.. since I don't have director 7 for Windows.. *yet*.

Sorry!
-steve

>>>>> "Steve" == Steve Spicklemire <steve at acer.spvi.com> writes:

    Steve> I've also been playing with teaching python to 'children'
    Steve> (well..  middle schoolers anyway... don't tell *them* they
    Steve> are children. ;->) I've had some luck with my little PyBots
    Steve> toy...  it needs some work, when I get more time.. but it's
    Steve> enough to teach some simple stuff...




From piet at cs.uu.nl  Tue Nov 16 04:45:02 1999
From: piet at cs.uu.nl (piet at cs.uu.nl)
Date: 16 Nov 1999 10:45:02 +0100
Subject: Need help dealing with spaces in NT paths.
References: <80pr76$jq5$1@nnrp1.deja.com> <11f733ec.96958819@usw-ex0101-007.remarq.com>
Message-ID: <uogcu93dt.fsf@cs.uu.nl>

>>>>> wheineman <wheinemanNOwhSPAM at uconect.net.invalid> (W) writes:

W> Put quotes around the complete path. For example:
W> myExec = "C:\\Program Files\\Trash Can\\compactor.exe"

You probably mean:
myExec = "\"C:\\Program Files\\Trash Can\\compactor.exe\""

or easier:
myExec = r'"C:\Program Files\Trash Can\compactor.exe"'
-- 
Piet van Oostrum <piet at cs.uu.nl>
URL: http://www.cs.uu.nl/~piet [PGP]
Private email: Piet.van.Oostrum at gironet.nl



From mwh21 at cam.ac.uk  Tue Nov 16 15:40:55 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 16 Nov 1999 20:40:55 +0000
Subject: problems with __getattr__ & __setattr__
References: <3F8BA119B48AD11197470008C7A493CB98B67D@HQOBGEX1>
Message-ID: <m3ogcu186g.fsf@atrus.jesus.cam.ac.uk>

ScherBi at BAM.com writes:

> I am having trouble using __getattr__ and __setattr__ in some classes.  I've
> boiled it down to the code below.
> It dumps core under 1.5.2 on Linux or WinNT.
> 
> I think maybe it's looping, if so, how does one go about doing this? 
> (I'm assuming it's clear enough what I'm trying to do.)
> 
> If I comment out the __gettattr__ operation, it doesn't dump core.  Instead
> I get an AttributeError: eggs raised at the 'self.eggs[key] = value' call in
> __setattr__.  This is what leads me to belive there's something circular
> going on here.
> 
> 
> Thanks,
>  Bill
> 
> -------------------------------------------------------------
> 
> class spam:
> 
>     def __init__(self):
>         self.eggs = {}

Hint: what does this call?
     
>     def __getattr__(self, key):
>         try:
>             return self.eggs[key]
>         except:
>             raise
>         
>     def __setattr__(self, key, value):

It calls this!

>         try:
>             self.eggs[key] = value

And this calls __getattr__ which looks for self.eggs, which calls
__getattr__, which looks for self.eggs, which calls __getattr__, which
looks for self.eggs, which calls __getattr__, which looks for
self.eggs, which calls __getattr__, which looks for self.eggs, which
calls __getattr__, 
Segmentation fault (core dumped)

>             return 0
>         except:
>             raise
> 
> 
> 
> if __name__ == '__main__':
> 
>     s = spam()
>     print 'instance of spam created.' 
> 

try

    def __init__(self):
        self.__dict__['eggs'] = {}

instead; also 

     def __getattr__(self, key):
         try:
             return self.__dict__["eggs"][key]
         except:
             raise

is probably safer. What are the try/catch for? They aren't doing
anything here!

__getattr__-and-__setattr__-are-so-much-fun-we-don't-mind-the-
performance-hit-ly y'rs - Michael



From leonar_n at epita.fr  Mon Nov 29 09:17:09 1999
From: leonar_n at epita.fr (leon)
Date: Mon, 29 Nov 1999 15:17:09 +0100
Subject: Unbuffered stdout
Message-ID: <81u1k2$27qkk$1@oceanite.cybercable.fr>

Hi there !

I'm developping an IRC like software which must work on the Web. I can't use
Java,
and I'm using Python to program the CGI.
But, I have a big problem : I don't know how to make the stdout unbuffered.
I tried to make some
    sys.stdout.write("test")
    sys.stdout.flush()
but it don't work.

So any idea ??


Leon.





From usenet-mm-yyyy at julesallen.com  Sun Nov 14 11:22:33 1999
From: usenet-mm-yyyy at julesallen.com (Jules Allen)
Date: Sun, 14 Nov 1999 16:22:33 GMT
Subject: Proposal: Python Info Collective
Message-ID: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net>

One of the main problems I've had with understanding Python's depth
is finding information and modules. c.l.py.announce is a good way
to keep track of new things but once that info scrolls from the
news server, it's pretty much gone. Sure, one can use deja.com (and
clones) but it's a chore.

On the module front, I had to update some Perl forms a while ago
and marveled at how easy it is to use CPAN. I'm not having a go at
the Trove development effort but unless I'm mistaken, it seems to
be stalled. Could  this be harming Python's wider adoption? On a
personal level, I can't tell if I should start a module to complete
a task or if is there already a module that fits.

Seasoned vererans and PyGods instantly know where everything is.
They politely and graciously seem to answer the same questions
about where to find tutorials, HTMLgen examples, PyApache's location
over and over again.

So what I'd like to propose is a yahoo/freshmeat flavour site tuned
for all things Pythonish. I'd say that Zope is the obvious candidate
for the development vehicle as it makes delegation of sub-sections
blissfully easy. If this project does fly, it's a strictly
non-commercial effort and Guido has suggested that subdomain.python.org
might be possible. Seems like the best place to host it to me.

I'm no library scientist or PyGod but I'll gladly organize, code
and do the bits nobody else wants to do. Rather than pollute c.l.py
with this discussion, I've created an eGroup at:

	http://www.egroups.com/list/pyinfo/

If you have any interest, ideas or flames, please sign up and let
fly!

Cheers,
Jules

PS I'll be at IPC8 and if you'd like to do the face to face thing
and chat about this, drop me a line.

--
If you want to email me directly, plug the current month and
year into my email address.




From cfelling at iae.nl  Sat Nov 27 17:12:57 1999
From: cfelling at iae.nl (Carel Fellinger)
Date: 27 Nov 1999 23:12:57 +0100
Subject: Newbie: raw_input and other input methods
References: <19991125.1481400@adam-aurag.penguinpowered.com> <81kfjc$1vg$1@vvs.superst.iae.nl> <383EC1A5.99A4FAD5@lemburg.com> <81pk1m$2i6$1@vvs.superst.iae.nl>
Message-ID: <81pl19$2lh$1@vvs.superst.iae.nl>

Carel Fellinger <cfelling at iae.nl> wrote:

I was to quick, or actually I changed something during posting.

> class Var:
>     '''(environment) variable to be evaluated during prompting'''
>     def __init__(self, var, dict=os.environ):
> 	self.var, self.dict = var, dict

>     def __str__(self):
> 	return self.dict[self.var]

this should read:

        return str(self.dict[self.var])

and now we can do

> if __name__ == '__main__':
      ...
     prompt.nr = Var('nr', locals())
     nr = 0
     while 1:
 	nr = nr + 1
 	print raw_input(prompt)

-- 
groetjes, carel



From skaller at maxtal.com.au  Thu Nov 25 18:26:15 1999
From: skaller at maxtal.com.au (skaller)
Date: Fri, 26 Nov 1999 10:26:15 +1100
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <8666ysfta9.fsf@ix.netcom.com>
Message-ID: <383DC597.1ED53168@maxtal.com.au>

Brad Knotwell wrote:
> In general, the only thing I've found annoying about Python are the scoping
> rules:
> 
> >>> def fact(n):
> ..     def fact_iter(n,result):
> ..       if not n:  return result
> ..       else:      fact_iter(n-1,n*result)
> ..     return fact_iter(n,1)
> ..
> >>> fact(3)
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
>   File "<stdin>", line 5, in fact
>   File "<stdin>", line 4, in fact_iter
> NameError: fact_iter

There's an error on your code (you have to 'return' in line 4).
After that correction, running on Viperi:

[root at ruby] ~/viper>./viperi
Viperi 2.0.1
Copyright Maxtal P/L, John Skaller, Australia, 1999

>>>def fact(n):
...   def fact_iter(n,result):
...     if not n:  return result
...     else:     return fact_iter(n-1,n*result)
...   return fact_iter(n,1)
...
>>>print fact(3)
...
6 
>>>print fact(7)
...
5040 

-- 
John Skaller, mailto:skaller at maxtal.com.au
10/1 Toxteth Rd Glebe NSW 2037 Australia
homepage: http://www.maxtal.com.au/~skaller
voice: 61-2-9660-0850



From m.faassen at vet.uu.nl  Mon Nov 15 06:22:32 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 15 Nov 1999 11:22:32 GMT
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net>
Message-ID: <80oqdo$ocq$1@newshost.accu.uu.nl>

Jules Allen <usenet-mm-yyyy at julesallen.com> wrote:
[snip]
> So what I'd like to propose is a yahoo/freshmeat flavour site tuned
> for all things Pythonish. I'd say that Zope is the obvious candidate
> for the development vehicle as it makes delegation of sub-sections
> blissfully easy. If this project does fly, it's a strictly
> non-commercial effort and Guido has suggested that subdomain.python.org
> might be possible. Seems like the best place to host it to me.

I think AppWatch is a good thing to look at then. It's a freshmeat style
site built on top of Zope. http://www.appwatch.com
So you should ask 'm for their sources. :)

> I'm no library scientist or PyGod but I'll gladly organize, code
> and do the bits nobody else wants to do. Rather than pollute c.l.py
> with this discussion, I've created an eGroup at:
[snip]

Hey, I thought comp.lang.python was for this kind of discussion. :)

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?



From news at dorb.com  Thu Nov 11 11:04:30 1999
From: news at dorb.com (Darrell)
Date: Thu, 11 Nov 1999 11:04:30 -0500
Subject: ASCII delimited files
References: <942231502.1914033961@news.interbusiness.it> <80ekhb$1qrh$1@hub.org>
Message-ID: <HMBW3.5644$JD1.378386@typhoon.nyroc.rr.com>

D'Arcy J.M. Cain wrote:
> I use delimited.py which works pretty good.

It sure does !
It ran 3.6 meg in 11.4 sec and my 'C' version ran in 1.75 sec
For most cases delimited.py would be fast enough.

The test case for delimited.py also pointed out a bug in my stuff.

--
--Darrell






From prestonlanders at my-deja.com  Tue Nov 30 18:50:53 1999
From: prestonlanders at my-deja.com (Preston Landers)
Date: Tue, 30 Nov 1999 23:50:53 GMT
Subject: how to do *simple* syntax checking?
Message-ID: <821nss$99v$1@nnrp1.deja.com>

 Okay, I know there was an earlier thread regarding pre-distribution
syntax checking (and the virtual impossibility of same).

 I understand Python is a dynamicly typed language, and I have no real
need to be able to find previously undeclared names in code before I
run it.

 However, it would be enormously helpful for me to be able to do
*simple* checks on Python code at make/'make install' time.  Like
catching whitespace/indentation problems, missing semicolons, and the
like.

 I maintain a large set of Python programs, some of which are accessed
through a CGI interface.  I develop some code, run make install,
restart my web server (not always though) and then load up my URL.
Only to find "500 Internal Server Error."

 Usually, if it's not a namespace issue, it's something silly like a
missing semicolon.  But I have to track through the server logs to find
out what went wrong.

 I thought that Python's compile() function would be what I want,
because I thought that the SyntaxError exception was thrown when making
bytecode.  But that's not the case, unless I was notably un-thorough in
my investigation.

 Basically, what I want is to do something like this for every Python
program in my makefile:

try:
   simple_syntax_check("myfile.py")
except:
   # file is no good, abort make process

# ... continue with make

 Am I just dreaming? Is this unrealistic?  Or is this kind of thing
fully availible and documented and I am just blind?

thanks,

---Preston

--
|| Preston Landers <prestonlanders at my-deja.com> ||


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



From fredrik at pythonware.com  Fri Nov 19 04:32:56 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 19 Nov 1999 10:32:56 +0100
Subject: Beginning programming with Python
References: <sng0ychgwt.fsf@motorola.com> <382B3A53.5816B2@callware.com> <382B3D79.74CD3E0F@email.sps.mot.com> <80p9ds$ppo$1@newshost.accu.uu.nl> <asteroidal-80psfo/INN-2.2.1/bat@broadway.news.is-europe.net> <38317787.19AFDAF5@compaq.com> <812rg4$b6$1@essle.valdyas.org>
Message-ID: <00ef01bf3271$10a8bb90$f29b12c2@secret.pythonware.com>

Boudewijn Rempt <boud at rempt.xs4all.nl> wrote:
> I tend to forget the colons after function and class definitions and
> everywhere else and the double == in if's - mainly because, for the life
> of Brian, I can't see why they are there. It seems to me the language
> would be just as easily parseable and readable without this bit of
> interpunction.

guess you weren't part of the CWI design team -- they
did test this on humans, you know, and discovered that
the language wasn't as easily parseable and readable
without the colons...

</F>

    "We did requirements and task analysis, iterative
    design, and user testing. You'd almost think pro-
    gramming languages were an interface between
    people and computers." -- Steven Pemberton




From phd at phd.russ.ru  Wed Nov  3 03:23:11 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Wed, 3 Nov 1999 08:23:11 +0000 (GMT)
Subject: Template for CGI
In-Reply-To: <7vn0sk$7cp$1@news.fsu.edu>
Message-ID: <Pine.LNX.4.20.9911030821220.30064-100000@fep132.fep.ru>

On Tue, 2 Nov 1999, Glenn Kidd wrote:
> I am a newbie to Python and I have a question.  I want a CGI script that
> will read a file template and replace certain words with variables.  Ii
> tried putting the variable names in the file but it is read as palin text.
> Is there a way to embed variables into a teplate or a replace function in
> Python?  Thank you for any help you give.

   Go to www.zope.org and download ZTemplates. Learning ZTemplates is
somewhat hard (not much, really), but it worth learning. After ZTemplates
you will NEVER need anything else! :)
   After you download it, I can send you some examples how to use it.

Oleg.
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From scott at hk.super.net  Wed Nov  3 06:42:05 1999
From: scott at hk.super.net (Scott Deerwester)
Date: Wed, 03 Nov 1999 11:42:05 GMT
Subject: JavaCard and JPython?
Message-ID: <19991103.11420500@kahua.transpacific.com.hk>

Is there any thought of implementing the JavaCard API using Jpython? 
The prospect of being able to program Smartcards in Python is very 
appealing indeed!







From andy at robanal.demon.co.uk  Thu Nov  4 16:23:06 1999
From: andy at robanal.demon.co.uk (Andy Robinson)
Date: Thu, 04 Nov 1999 21:23:06 GMT
Subject: Drawing (directed) graphs
References: <001701bf2540$bd4d07e0$3acbd9c2@peridot.optichrome.com>
Message-ID: <3823f33e.11772788@news.demon.co.uk>

"Adrian Eyre" <a.eyre at optichrome.com> wrote:

>> Does anyone know of Python code for drawing graphs of the
>> graph-theoretic type? Ideally, I'd like to be able to interactively
>> add nodes and edges etc and add data to structures associated with the
>> nodes and edges.
>
>Depends if you're looking for a proper graphics display, or an ASCII
>representation. A full graphics version would not be platform
>independent, and would require a graphics library.

Good thing there is a platform independent Python graphics library!
PIDDLE (Plug In Drawing, Does little Else) at
www.strout.net/python/piddle
There is a discussion list at www.egroups.com/pythonpiddle/

PIDDLE is pretty stable, and we hope we'll build up a bunch of such
applets on top of it over time.

BTW, I'm in charge of the PDF back end for PIDDLE, and would certainly
have uses for such a routine.  One demo I've always wanted is to look
at a Python program and draw a module dependency graph (as well as all
the module docstrings)....

Regards,

Andy



From robin at jessikat.demon.co.uk  Thu Nov  4 08:13:33 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Thu, 4 Nov 1999 13:13:33 +0000
Subject: Comparison between Python and "Ruby"
References: <1270552684-31022246@hypernet.com>
 <87n1svugfh.fsf@ev.netlab.co.jp> <7vrtch$nro$1@newshost.accu.uu.nl>
Message-ID: <Y9gbyBA9ZYI4Ew5R@jessikat.demon.co.uk>

I'm fairly sure all of these languages are turing complete and if so
they're equivalent at the semantic level. To defeat TC once and for all
let's implement a PyPerl module etc etc etc. Greg is at least doing some
interesting comparisons all this n ways to babylon and back stuff or how
many T's H's etc is a bit of a waste.
-- 
Robin Becker



From MSteed at altiris.com  Tue Nov 30 18:36:23 1999
From: MSteed at altiris.com (Mike Steed)
Date: Tue, 30 Nov 1999 16:36:23 -0700
Subject: What's the canonical vi setting for Python indentation
Message-ID: <65118AEEFF5AD3118E8300508B124877073CA2@webmail.altiris.com>

> From: Chris WRIGHT [mailto:caw at cs.mu.oz.au]
> Sent: Tuesday, November 30, 1999 4:19 PM
> To: python-list at python.org
> Subject: What's the canonical vi setting for Python indentation
> 
> 
> I've tried to find out the "OK" indentation settings for 
> python with vi,
> and a troll through deja-news found some hints (like don't do indent =
> 4). Yet I seem to also remember some advice to the contrary?? I want
> indentation to be 4 characters, and readable by/ acceptable to emacs
> python-mode, and to pass tabnanny...

I haven't tested acceptability to emacs or tabnanny, but since sane software
assumes a tabstop of 8, I use (in vim) ':set ts=8 sts=4 sw=4'.  This makes
tabs feel like four spaces while I'm editing (convenient), but uses actual
tab characters only for groups of 8 spaces.

--
M.



From fredrik at pythonware.com  Wed Nov 24 03:55:44 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 24 Nov 1999 09:55:44 +0100
Subject: dir() and propertly list NEWBIE
References: <81eorl$b60$1@nnrp1.deja.com>
Message-ID: <00ce01bf3659$b5cddce0$f29b12c2@secret.pythonware.com>

<tiddlerdeja at my-deja.com> wrote:
> Is it possible to get the functions of an object using dir()? Rather
> that looking them up in a book/API.
>
> Similarly, If I can get an objects functions, is it possible then get a
> list of the possible parameters for that function?

the attached module contains all you need.

</F>

from the eff-bot archives:

#
# describeFunction.py
# $Id$
#
# utilities to describe functions, methods, and classes
#
# history:
# 96-10-27 fl     created
# 98-02-24 fl     added code to handle unpacked arguments
# 99-11-24 fl     sample code used bogus argument order
#
# notes:
# This has been tested with Python 1.4 and 1.5.  The code and
# function object attributes might change in future versions of
# Python.
#
# written by fredrik lundh.  last updated november 1999.
#
# fredrik at pythonware.com
# http://www.pythonware.com
#

import string

# --------------------------------------------------------------------
# code object attributes
# --------------------------------------------------------------------
# co_argcount   INT
# co_nlocals    INT
# co_flags      INT
#       CO_OPTIMIZED
#       CO_NEWLOCALS
#       CO_VARARGS
#       CO_VARKEYWORDS
# co_code       OBJECT
# co_consts     OBJECT
# co_names      OBJECT
# co_varnames   OBJECT
# co_filename   OBJECT
# co_name       OBJECT

# --------------------------------------------------------------------
# function object attributes
# --------------------------------------------------------------------
# func_code     OBJECT
# func_globals  OBJECT
# func_name     OBJECT (__name__)
# func_defaults OBJECT
# func_doc      OBJECT (__doc__)

# copied from Python header file
CO_OPTIMIZED = 0x0001
CO_NEWLOCALS = 0x0002
CO_VARARGS = 0x0004
CO_VARKEYWORDS = 0x0008

def _describe(func, name = None):
    # get argument list

    code = func.func_code

    n = code.co_argcount
    a = list(code.co_varnames[:n])
    p = 0
    for i in range(n):
        # anonymous arguments
        from dis import opname, HAVE_ARGUMENT    
        c = code.co_code
        if not a[i] or a[i][0] == ".":
            vars = []
            while p < len(c):
                v = ord(c[p])
                if v >= HAVE_ARGUMENT:
                    s, v = opname[v], ord(c[p+1]) + ord(c[p+2])*256
                    p = p + 3
                    if s == "UNPACK_TUPLE":
                        count = v
                    elif s == "STORE_FAST":
                        vars.append(code.co_varnames[v])
                        if len(vars) >= count:
                            break
                else:
                    p = p + 1
            if vars:
                a[i] = "(" + string.join(vars, ", ") + ")"
    if func.func_defaults:
        # defaults
        i = n - len(func.func_defaults)
        for d in func.func_defaults:
            a[i] = (a[i], d)
            i = i + 1
    if code.co_flags & CO_VARARGS:
        # extra arguments
        a.append("*"+code.co_varnames[n])
        n = n + 1
    if code.co_flags & CO_VARKEYWORDS:
        # extra keyword arguments
        a.append("**"+code.co_varnames[n])
        n = n + 1
    return a

def describe(func, name = None):
    "Return the function or method declaration as a string"

    # argument list
    a = _describe(func)
    args = []
    for arg in a:
        if type(arg) is type(""):
            args.append(arg)
        else:
            args.append("%s=%s" % (arg[0], repr(arg[1])))
    args = string.join(args, ", ")

    # function name
    if not name:
        name = func.func_name
        if name == "<lambda>":
            return "lambda %s" % args
    return "%s(%s)" % (name, args)


def __getmethods(c, m):
    for k, v in c.__dict__.items():
        if type(v) is type(__getmethods): # and k[0] != "_":
            if not m.has_key(k):
                m[k] = describe(v, k), c.__name__
    for c in c.__bases__:
        __getmethods(c, m)


def describe_class(cls):
    "Return a dictionary describing all methods available in a class"

    m = {}
    __getmethods(cls, m)
    return m


def describe_instance(self):
    "Return a dictionary describing all methods available in an instance"

    return describe_class(self.__class__)

#
# --------------------------------------------------------------------

if __name__ == "__main__":

    def foo(a, b=1, *c, **d):
        e = a + b + c
        f = None

    bar = lambda a: 0

    # based on an example from Duncan Booth
    def baz(a, (b, c, d), e, (f, g) = ('foo','bar')):
        pass

    print describe(foo)
    print describe(bar)
    print describe(baz)

# --- end of module ---




From garry at research.att.com  Tue Nov  2 13:28:02 1999
From: garry at research.att.com (Garrett G. Hodgson)
Date: Tue, 2 Nov 1999 18:28:02 GMT
Subject: tkinter grid() question
Message-ID: <381F2D32.9E3A15CF@research.att.com>

in the code below, i create a Text entry area with a Scrollbar.
i want this pair to always use all of the available space.
but when i resize the window, the stay the same size, neither
expanding or shrinking.

what am i doing wrong?

thanks

--------------------
from Tkinter import *

class Bug(Frame):
	def __init__( self, parent ):
		Frame.__init__( self, parent )
		self.pack( fill=BOTH )

		scrollbar = Scrollbar( self, orient=VERTICAL )
		self.body = Text( self )

		self.body.config( yscrollcommand=scrollbar.set )
		self.body.grid( row=2, column=0, rowspan=10, columnspan=10, sticky=N+S
)
		scrollbar.config( command=self.body.yview )
		scrollbar.grid( row=2, column=10, rowspan=10, sticky=N+S )

if __name__ == '__main__':
	root = Tk()
	Bug( root )
	root.mainloop()

-- 
Garry Hodgson			"Hey, mister, can ya tell me,
garry at sage.att.com		where a man might find a bed?"
Software Innovation Services	He just grinned and shook my hand,
AT&T Labs			"No", was all he said.



From kantel at mpiwg-berlin.mpg.de  Tue Nov  9 04:52:09 1999
From: kantel at mpiwg-berlin.mpg.de (Jörg Kantel)
Date: Tue, 09 Nov 1999 10:52:09 +0100
Subject: Graphics.
References: <3826E949.BBA1BAFD@imf.au.dk>
Message-ID: <091119991052098844%kantel@mpiwg-berlin.mpg.de>

In article <3826E949.BBA1BAFD at imf.au.dk>, Emil Hedevang Lohse
<emil at imf.au.dk> wrote:

> How do I draw some graphics on a screen using Python running on
> HP-UX
> MacOS
> Windows
> 
> ???
> 

If you want to use Tkinter and maybe you can read German than I've a
nice example at my website at
 
<http://www.mpiwg-berlin.mpg.de/staff/kantel/pythonmania/chapter04.html>


HTH
J"org Kantel



From jam at quark.emich.edu  Tue Nov 30 17:38:37 1999
From: jam at quark.emich.edu (Jeff)
Date: Tue, 30 Nov 1999 17:38:37 -0500
Subject: Looking for sshlib.py
In-Reply-To: <tBX04.11350$VA6.54975@news-server.bigpond.net.au>
References: <iwI04.9621$VA6.47863@news-server.bigpond.net.au> <19991130080326.A1376@quark.emich.edu> <tBX04.11350$VA6.54975@news-server.bigpond.net.au>
Message-ID: <19991130173837.A2365@quark.emich.edu>

On Tue, Nov 30, 1999 at 09:47:05PM +0000, John Leach wrote:
> I'm looking to write a Python frontend for character-based applications
> accessed via the internet.
> telnet works well but I'd like something with more security (telnet sends
> its password unencrypted I believe).
> I'll try calling ssh from within Python and see if that helps.
> Thanks for your reply,
> John Leach
> http://www.bosware.com.au
> 

you are correct in your statement that telnet sends it's passwords in the
clear. ssh has clients designed to replace 'rsh' and 'rcp'. if you could
implement your application with rsh first, ssh will work nicely as a secure
replacement. in other cases I would suggest using a different piece of
software.. perhaps kerberos.

kerberos homepage: http://web.mit.edu/kerberos/www/
ssh homepage: http://www.ssh.org/ or http://www.ssh.fi/

regards,
J
-- 
|| visit gfd <http://quark.emich.edu/>
|| psa member #293 <http://www.python.org/> 
|| New Image Systems & Services, Inc. <http://www.newimage.com/>



From thomas.heller at ion-tof.com  Thu Nov  4 02:51:00 1999
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Thu, 4 Nov 1999 08:51:00 +0100
Subject: Embedding and threads
Message-ID: <00b501bf2699$567dccc0$4500a8c0@thomasnotebook>

While embedding python in a DLL I'm writing (we are talking Windows here),
I encountered the following confusion:

Threads are created in the application from C. From reading the docs I
understood that I must 'boot the thread' into existance by creating a thread
state data structure (with PyThreadState_New(), passing the
PyInterpreterState *interp pointer which I saved from the main thread),
then acquire the interpreter lock, and finally store the thread state
pointer before using the Python/C API from this thread.

I do this in the DllMain() function in the DLL_THREAD_ATTACH branch.

This all works very well, but what happens if I also create threads
from python? These threads will be initialized by python, so
I assume the thread state data structure is already created.

How do I ask python if the thread state already exists for a given thread?

Thomas Heller
ION-TOF GmbH




From a.eyre at optichrome.com  Wed Nov  3 12:17:56 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Wed, 3 Nov 1999 17:17:56 -0000
Subject: __LINE__?
In-Reply-To: <816010E2456BD111A48700805FBBE2EEFDF750@ex-quebec-u1.baan.com>
Message-ID: <002a01bf261f$5eae0dd0$3acbd9c2@peridot.optichrome.com>

> Is there an equivalent of the C __LINE__ macro in Python to get the
> currentline number?

import sys

def what_line_am_i_on():
    try:
        raise "Hack"
    except:
        return sys.exc_info()[2].tb_frame.f_back.f_lineno

print "This is line " + str(what_line_am_i_on())

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From tim_one at email.msn.com  Fri Nov 12 00:04:47 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Fri, 12 Nov 1999 00:04:47 -0500
Subject: print "foo", without a space
In-Reply-To: <382AD3BB.69899CBD@helsinki.xx>
Message-ID: <000b01bf2ccb$70ba3d20$fd2d153f@tim>

[Mikael Johansson]
> Is there another way of getting print to not make a newline than
> using:
>     print "foo",  ?

No.

> This because the comma always produces a space which necessary
> isn't that nice ...

That's a different issue than newlines.  In small doses, this devious trick
is quite pleasant:

>>> import sys
>>> def no_space_before(x):
        sys.stdout.softspace = 0
        return x

>>> ns = no_space_before
>>> print 1, ns(2), 3, ns(4), 5
12 34 5
>>>

the-great-thing-about-python-is-it's-so-obvious<wink>-ly y'rs  - tim





From gerrit.holl at pobox.com  Thu Nov 18 15:39:25 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Thu, 18 Nov 1999 21:39:25 +0100
Subject: urllib.getsize()
Message-ID: <19991118213925.A10221@optiplex.palga.uucp>

Hello,

I've written a function to get the size of a HTTP/FTP url, is it good enough
to be put in urllib.py?

def getsize(url):
        import urlparse
        urltup = urlparse.urlparse(url)

        protocol = urltup[0]
        host = urltup[1]
        file = urltup[2]

        if protocol == HTTP:
                import httplib
                http = httplib.HTTP(host) # www.nl.linux.org:80 mag ook
                http.putrequest("HEAD", file)
                http.putheader("Accept", "*/*")
                http.endheaders()
                return http.getreply()[2].get("Content-Length")
        elif protocol == FTP:
                import ftplib
                ftp = ftplib.FTP(host) # ftp.nl.linux.org:1234 mag ook
                ftp.login()
                return ftp.size(file)
	else:
		return -1

Maybe there should be some more error checking...

regards,
Gerrit.

-- 
"I'd crawl over an acre of 'Visual This++' and 'Integrated Development
That' to get to gcc, Emacs, and gdb.  Thank you."
(By Vance Petree, Virginia Power)



From tekhir at yahoo.com  Sat Nov 13 22:34:10 1999
From: tekhir at yahoo.com (Jeremy Misavage)
Date: Sat, 13 Nov 1999 22:34:10 -0500
Subject: Python and sending emails with it
Message-ID: <382E2DB2.A00B9109@nv2.cc.va.us>

I'm kinda new to python and I'm trying to write a small program that will take
in form data, save it in a db, and then send an email to a person.

The program processes a users request for a "equipment request."  It gets
loggged in a db (already done) and sends an email to the person in change of the
equipment.  I need help on the email part.  Does some out there have some code
that will taken data and send it to a person.  It shouldn't need a username and
password, but if it does I'll live with that.



From stannous at employees.org  Thu Nov  4 15:03:25 1999
From: stannous at employees.org (Sam Tannous)
Date: Thu, 04 Nov 1999 15:03:25 -0500
Subject: Questions on importing and exec command.
Message-ID: <3821E68D.D785F4F3@employees.org>

I'm writing a program that starts a separate thread
where the user can type python commands (which
are executed with the exec command).  This saves me
the trouble of having to write a parser, etc... I realize
the hazards in doing this ;-)

I'm having several problems:

1.  A minor problem is command retrieval (or
    command history) in the console. If someone types a fairly
    long command, I'm not sure how to make the
    up and down arrows work correctly (as in tcsh)
    from the console I've created.   Some key mapping
    within the program?

2.  I want to allow the user (once she's in the console
    I've provided) to "import" or otherwise run python
    code fragments that have access to the functions I've
    provided globally. Typing the function name in the
    console seems to work just fine.  But "import"ing a small
    chunk of code doesn't seem to work (and I don't want
    to force users to type everything by hand).  The files
    I'm importing don't see the global functions I've declared
   (see the error in the sample session below).   I want to give the
    user complete freedom to write python code fragments and
    save them in files and then just run them from the console
    as easily as possible.

Here is the code:

------------  test2.py ---------
#!/usr/bin/python
import time,thread,string,sys

def main():
    print 'in main'
    thread.start_new(listener,('test',))
    myconsole()

def listener(test):
    global count
    count = 0
    while 1:
        print "in listener",test,count
        time.sleep(30)
        count = count + 1

def myconsole(*x):
    print 'Now in console'
    while 1:
        command = raw_input('console > ')
        command = string.strip(command)

        try:
            exec command in globals()
        except:
            print sys.exc_type,sys.exc_value

def testcall(message='default'):
    print "in testcall",message

def startit(hold_time=10):
    thread.start_new(do_stuff,(hold_time,))

def do_stuff(hold_time=10):
    print 'started do_stuff  hold_time =',hold_time
    time.sleep(hold_time)
    print 'ended do_stuff'

def do_nothing(*x):
    pass


count = 0
my_dict = {'01':('test1',startit),'07':('test2',do_nothing)}

main()
------------------------


-------- call1.py -------
for i in range(5,9):
    print i
    startit(hold_time=i)
---------------------------

Here's a session where I get the error message
when calling function startit from an import statement.
(stuff the program prints has a '****' in front of it)
-------------------------
$test2.py
****in main
****Now in console
****in listener test 0
console >
console >
console > testcall()              #here I call a function without
problems
****in testcall default
console > ****in listener test 1
****in listener test 2
****in listener test 3
****in listener test 4
****in listener test 5
****in listener test 6
****in listener test 7

console > ****in listener test 8

console > count = 666            #changing a global variable is no
problem
console >
console > testcall()
****in testcall default
console > print my_dict          #looking at global variables is no
problem
{'01': ('test1', <function startit at 27bba0>), '07': ('test2',
<function do_nothing at 293688>)}
console > ****in listener test 667

console > startit()
console > ****started do_stuff  hold_time = 10
****ended do_stuff

console > start****in listener test 668
^R

console > startit(hold_time=3)
console > ****started do_stuff  hold_time = 3

console > ****ended do_stuff

console > import call1           # importing a code fragment causes
problems
5
exceptions.NameError startit
---------------------------

Is there a trick to importing code fragments from this python session
I've created?

Thanks,
Sam




From guido at CNRI.Reston.VA.US  Wed Nov  3 20:57:12 1999
From: guido at CNRI.Reston.VA.US (Guido van Rossum)
Date: Wed, 03 Nov 1999 20:57:12 -0500
Subject: __str__ vs. __repr__
In-Reply-To: Your message of "Wed, 03 Nov 1999 00:08:56 EST."
             <000401bf25b9$88ea57e0$2e2d153f@tim> 
References: <000401bf25b9$88ea57e0$2e2d153f@tim> 
Message-ID: <199911040157.UAA11908@eric.cnri.reston.va.us>

[Tim Peters, arguing against the current design of str() and repr()]

> [Randall Hopper wonders about str-vs-repr, Tim explains that lists ask
>  their elements to produce repr() even if the list itself was passed
>  to str()
> ]
> 
> [Guido]
> > Actually, Python's internals *do* follow the convention.  Dicts and
> > lists don't define __str__(), so str() for them defaults to
> > __repr__(), and then of course repr() is used for the items.  It may
> > be confusing, it may not be what you want, but it *is* consistent ;-(
> 
> Not to mention undocumented <wink>.
> 
> [Randall]
> > For consistency, would it make sense to change this for Python 1.5.3 (that
> > is, have sequence and dict types pass 'str-vs-repr'ness down)?
> 
> [Guido]
> > This has been asked a few times before (and usually it's been reported
> > as a bug, which it isn't -- see above).  I happened to see this post
> > and it made me delve deep into my intuition trying to figure out why I
> > don't like propagating str() down container items.
> >
> > Here's what I think it is.  There's no reason why an object's str()
> > should be particularly suited to being included in list syntax.
> 
> This seems much more a consequence of the current design than an argument in
> favor of it.  That is, had Python been designed so that the builtin
> container types did "pass down" str-vs-repr'ness, an object's str() would
> have every reason to produce a string suited to being etc.
> 
> > For example, I could have a list containing the following items:
> >
> > 	1		# an integer
> > 	'1'		# a string (of an integer literal)
> > 	'2, 3'		# a string containing a comma and a space
> > 	'], ['		# a string containing list delimiters
> >
> > Under the proposed rules, this list would print as:
> >
> > 	[1, 1, 2, 3], []
> >
> > I would find this confusing
> 
> Me too, but I find the current design more rigidly consistent than useful
> (see below).  In a world where containers passed str() down, a container's
> str() would presumably be responsible for adding disambiguating delimeters
> to element str() results when needed (the container knows its own output
> syntax, and can examine the strings produced by its elements -- not rigidly
> consistent, but useful <wink>).

Hm...  What kind of things would you expect e.g. the list str() to do
to its item str()s?  Put backslashes before commas?

> > and I worry that it could be used to fool the user.
> 
> People can already define __repr__ to return anything whatsoever; the
> reports of people getting fooled by this are conspicuous by absence <wink>.
> 
> Here's something wholly typical of what I dislike:
> 
> >>> from Rational import Rat, Format
> >>> Rat.set_df(Format.Format(mode=Format.FIXED, prec=3, use_tag=0))
> Format(mode=Format.MIXED, prec=8, base=10, round=Round.NEAREST_EVEN,
>        use_tag=1, use_letters=1)
> >>> one_tenth = Rat.Rat(.1)
> >>> one_tenth
> Rat(3602879701896397L, 36028797018963968L)
> >>> print one_tenth
> 0.100
> >>>
> 
> That is, in interactive mode, I'm forever using "print" because the default
> of applying repr() to raw expressions produces the output least useful in
> interactive hacking (I don't care about reproducing the object exactly from
> the string when typing a raw expression at the prompt!  The mental ratio of
> two giant integers isn't helpful here.).
> 
> Carry it one more step, and nothing simple suffices anymore:
> 
> >>> values = [one_tenth, one_tenth + 100]
> >>> values
> [Rat(3602879701896397L, 36028797018963968L),
>  Rat(3606482581598293197L, 36028797018963968L)]
> >>> print values
> [Rat(3602879701896397L, 36028797018963968L),
>  Rat(3606482581598293197L, 36028797018963968L)]
> >>>
> 
> So I'm forever typing this instead:
> 
> >>> map(str, values)
> ['0.100', '100.100']
> >>>
> 
> Throw a dict into it, and it's hopeless:
> 
> >>> recip = {one_tenth: 1/one_tenth, 1/one_tenth: one_tenth}
> >>> print recip
> {Rat(3602879701896397L, 36028797018963968L):
>      Rat(36028797018963968L, 3602879701896397L),
>  Rat(36028797018963968L, 3602879701896397L):
>      Rat(3602879701896397L, 36028797018963968L)}
> >>>
> 
> Having gone thru the same business in dozens of classes over the years, I
> find the current design simply unusable in interactive mode.  For a while I
> defined just __str__, bound __repr__ to that too, and added a .repr()
> *method* for the unusual cases in which I really needed a faithful string.
> But that frustrated other code that expected explict repr() calls, and/or
> the `` notation, to produce the long-winded version.  So that sucked too.
> 
> It's even an irritation sticking to builtin types; e.g., here assuming
> Latin-1 comes across intact:
> 
> >>> names = ["Fran?ois", "Tim"]
> >>> print names[0]
> Fran?ois
> >>> >>> print names
> ['Fran\347ois', 'Tim']
> >>>
> 
> That isn't helpful either -- it's frustrating.

These are all good points.

In a typical scenario which I wanted to avoid, a user has a variable
containing the string '1' but mistakenly believes that it contains the
integer 1.  (This happens a lot, e.g. it could be read from a file
containing numbers.)  The user tries various numeric operations on the
variable and they all raise exceptions.  The user is inexperienced and
doesn't understand what the exceptions are, but gets the idea to
display its value to see if something's wrong with it.  One of the
first things users learn is to use interactive Python as a power
calculator, so my hypothetical user just types the name of the
variable.  If this would use str() to format the value, the user is no
wiser, and perhaps more confused, since str('1') is the same as
str(1).  So I designed Python's read-eval-print loop to use repr()
instead of str(): when the user tries to display the variable, it will
show the string quotes which are a pretty good hint that it's a
string.  (Alternative scenario: the user has this problem and shows it
to a somewhat more experienced user, who displays the variable and
notes the problem from the output.)

Where have I gone wrong?  It seems that you are suggesting that the
read-eval-print loop (a.k.a. the >>> prompt or the interactive
interpreter) should use str(), not repr().  This would solve your
first example; if str() for lists, tuples and dictionaries were to
apply str() to their items, your second and third example would also
be solved.

We can then argue over what str() of a list L should return; one
extreme possibility would be to return string.join(map(str, L)); a
slightly less radical solution would be '[' + string.join(map(str, L),
', ') + ']'.  In the first case, your last example would go like this:

>>> names
Fran?ois Tim
>>> 

while the choice would give

>>> names
[Fran?ois, Tim]
>>>

There may be other solutions -- e.g. in Tcl, a list is displayed as
the items separated by spaces, with the proviso that items containing
spaces are displayed inside (matching) curly braces; unmatched braces
are displayed using backslashes, guaranteeing that the output can be
parsed back into a list with the same value as the original.  (Hey!
That's the same as Python's rule!  So why does it work in Tcl?
Because variables only contain strings, and the equivalent of the Rat
class used above can't be coded in Tcl.)

The problem with displaying 'Fran?ois' has been mentioned to me
before.  (By the way, I have no idea how to *type* that.  I'm just
cutting and pasting it from Tim's message.)

There's another scenario I was trying to avoid.  This is probably
something that happened once too many times when I was young and
innocent, so I may be overracting.  Consider the following:

>>> f = open("somefile")
>>> a = f.readline()
>>> print a
%????1.3

>>> 

Now this example is relatively harmless.  Using repr(), I see that the
string contains a \r character that caused the cursor to back up to
the start of the line, overwriting what was already written:

>>> a
'%PDF-1.3\015%\342\343\317\323\015\012'
>>>

But the thing that some big bad program did to me long ago was more
like spit out several thousand garbage bytes which contained enough
escape sequences to lock up my terminal requiring me to powercycle and
log in again.  (The fact that the story refers to a terminal indicates
how long ago this was. :-)

So I vowed that *my* language would not (easily) let this happen by
accident, and the way I enforced that was by making sure that all
non-ASCII characters would be printed as octal escapes, unless you use
the print statement.

It's a separate dilemma from the other examples.  My problem here is
that I hate to make assumptions about the character set in use.  How
do I know that '\237' is unprintable but '\241' is a printable
character?  How do I know that the latter is an upside-down
exclamation point?  Should I really assume stdout is capable of
displaying Latin-1?  Strictly, the str() function doesn't even know
that it's output is going to stdout.  I suppose I could use isprint(),
and then Fran?ois could use the locale module in his $PYTHONSTARTUP
file to make it do the right thing.  Is that good enough?  (I just
tried this briefly.  It seems that somehow the locale module doesn't
affect this?!?!

I still think that ['a', 'b'] should be displayed like that, and not
like [a, b].  I'm not sure what that to do about a dict of Rat()
items, except perhaps giving up on the fiction that __repr__() should
return something expression-like for user-defined classes...

Any suggestions?

--Guido van Rossum (home page: http://www.python.org/~guido/)



From gerrit.holl at pobox.com  Mon Nov  1 09:04:51 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Mon, 1 Nov 1999 15:04:51 +0100
Subject: Destroying files
In-Reply-To: <slrn81pmvn.1md.erno@fun77.kivikko.hoas.net>
References: <1270962155-6437791@hypernet.com> <slrn81pmvn.1md.erno@fun77.kivikko.hoas.net>
Message-ID: <19991101150451.A26174@optiplex.palga.uucp>

Erno Kuusela wrote:
> On Thu, 28 Oct 1999 21:38:09 -0400, Gordon McMillan <gmcm at hypernet.com>
> wrote:
> 
> writing zeros over the data does not make it unrecoverable.
> there are numerous companies that specialise in this kind
> of recovery.

You should do it three times.

> now, the python file objects seem to be missing a melt() method...
> perhaps this can be added for python 2.0? or perhaps a more suitable
> name would be self_destruct(), so it would not be tied to any single
> method.

.destroy() for writing zeros over it three times and unlink() afterwards.

regards,
Gerrit.
-- 
linuxgames.nl.linux.org	 All about games on Linux in Dutch (under construction).
www.nl.linux.org	       The Dutch resource for Dutch Linux documentation.
www.nl.linux.org/discoverb		    Learn foreign words and definitions.
www.nl.linux.org/~gerrit/asperger	 I'm an Asperger, go here for more info.
www.nl.linux.org/~gerrit			    About me and pictures of me.



From guido at CNRI.Reston.VA.US  Fri Nov  5 13:44:13 1999
From: guido at CNRI.Reston.VA.US (Guido van Rossum)
Date: Fri, 05 Nov 1999 13:44:13 -0500
Subject: Conference Registration is now Open
Message-ID: <199911051844.NAA17378@eric.cnri.reston.va.us>

On-line registration for the upcoming Python conference is now open.
New this year: we now accept on-line payment.

  http://www.python.org/workshops/2000-01/registration.html

Important dates:

  Jan 3: early bird registration + payment, hotel reservation
  Jan 24-27: The Conference!

See you all in Arlingon...  http://www.python.org/workshops/2000-01/

--Guido van Rossum (home page: http://www.python.org/~guido/)



From gerrit.holl at pobox.com  Fri Nov 19 07:53:59 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Fri, 19 Nov 1999 13:53:59 +0100
Subject: Beginning programming with Python
In-Reply-To: <383195C1.826B3342@iqsoft.hu>
References: <sng0ychgwt.fsf@motorola.com> <382B3A53.5816B2@callware.com> <382B3D79.74CD3E0F@email.sps.mot.com> <80p9ds$ppo$1@newshost.accu.uu.nl> <asteroidal-80psfo/INN-2.2.1/bat@broadway.news.is-europe.net> <38312D55.A36EB56D@iqsoft.hu> <antigorite-80rol2/INN-2.2.1/attentive@broadway.news.is-europe.net> <383195C1.826B3342@iqsoft.hu>
Message-ID: <19991119135359.A4306@optiplex.palga.uucp>

Nemeth Miklos wrote:
> Janos Blazi wrote:
> 
> I am neither a physicist nor a mathematician, I am a simple business software
> engineer. I like Python but I'd like to be convinced that Python is not only
> for scientific programming, but also for business programming. I'd like
> to know the general opinion of the "official" Python community on this

I won't claim to be the official python community, but there are many
business things done in Python. Electronic Arts is using Python for their
games. RedHat uses python for Linux installation. On the python site are many
more commercial projects done in Python.

-- 
"We all know Linux is great...it does infinite loops in 5 seconds."
(Linus Torvalds about the superiority of Linux on the Amterdam
Linux Symposium)



From paulb at infercor.no  Thu Nov 18 09:41:30 1999
From: paulb at infercor.no (Paul Boddie)
Date: Thu, 18 Nov 1999 15:41:30 +0100
Subject: Trouble with Bobo -- use Zope?
References: <80v5g8$5a6$1@mulga.cs.mu.OZ.AU>
Message-ID: <3834101A.FF39F64C@infercor.no>

Tim Bell wrote:
> 
>  3. Should I be moving to Zope and/or ZPublisher?  Is there a quick
>     guide to making the switch?  (I don't need all the fancy Zope stuff
>     at this stage.)

It really depends. Zope, when used with ZServer, is good at removing most of the
hassles associated with setting things up to work with the CGI mechanisms of the
average Web server, and provides a powerful framework for site management.
However, many people may not have the possibility of running their own server
and I haven't tried that hard to integrate Zope with a different Web server
using CGI.

For many people, I suspect that Zope is a bit "too much" when all they really
require is a nicer way to write server-side programs; this is what attracted me
to Bobo in the first place. Handling CGI inputs using the rather comprehensive
standard 'cgi' module gets very tedious and doesn't make for particularly nice
code. Indeed, it takes one's concentration away from the higher level parts of
an application because of the requirement to get right down to arguably
unnecessary detail.

I will admit that once the hassles of getting Bobo programs to work are over
(Zope products are also hard to get into to start with) it is a powerful tool
for developing applications at a higher level of sophistication than the usual
stand-alone CGI. ZPublisher and ZTemplates (or whatever DocumentTemplate is
called these days) should still provide you with Bobo's functionality.

As for ZPublisher, I moved onto Zope from ZPublisher with the idea of trying
something more ambitious out. Although Zope is very nice, I still feel that
there is a place for a simple object wrapper in the spirit of the original Bobo
software, or the ILU HTTP facilities, with minimal effort required in order to
get started. Personally, I don't really need persistent object support for
everything; especially if I am writing "wrappers" for command line tools, for
example.

Learning Zope is a fairly big effort, and can cause a fair amount of irritation
at times. The good news for beginners now is the effort being put in to improve
the documentation - not so long ago, the best way to learn certain features was
to read other people's source code and guess at its meaning.

Paul



From pinard at iro.umontreal.ca  Mon Nov  1 11:44:44 1999
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 01 Nov 1999 11:44:44 -0500
Subject: Produce .pyc without execution
In-Reply-To: "Adrian Eyre"'s message of "Mon, 1 Nov 1999 09:50:39 -0000"
References: <002001bf244e$8e187600$3acbd9c2@peridot.optichrome.com>
Message-ID: <oq904ii2kj.fsf@titan.progiciels-bpi.ca>

"Adrian Eyre" <a.eyre at optichrome.com> writes:

> > However, I could not decently propose this scheme for integration in
> > Automake, and that makes it a bit unsatisfactory.  It would be nicer,
> > for example, if `src/*.py' files were recompiled into the `src/' build
> > hierarchy from the `src/' distribution hierarchy, and then installed from
> > there.  But `compileall' wants to install compiled files next to sources
> > in the same directory, which is a bit inconvenient.  Any simple solution?

> This is the default behaviour for Python, and without modifying its source
> the .pyc will always be created in the same dir as the source. Can't
> you add something like:

> install-data-hook:
> 	@echo 'import compileall; compileall.compile_dir("$(pkgdatadir)")' \
> 	| $(PYTHON); \
> 	for pyc in $(pkgdatadir)/*.pyc; do mv $$pyc somewhere_else/; done

Not really.  There are many things to keep in mind for fitting well in
the Automake spirit, and these two are a bit more relevant, here:

* The distribution hierarchy (by opposition to build hierarchy) should be
  considered read-only, as if, say, it was already exploded in a CD-ROM; a
  mere, simple installation should not modify, create or delete files in it;
  this is only allowed in the build hierarchy.

* Many people do `make all' as one user, and `make install' as a different
  user (`root' or `bin', say), so ideally, all needed files should be
  prepared in the build hierarchy at `make all' time, `make install'
  should not create any more files in the build directory, once `make all'
  has been done.  The only files `make install' may create sit in the various
  installation directories.

Above, $(pkgdatadir) is an installation directory, so things are already in
their final place, no need to move them further once there.  The scheme I
used is acceptable when one wants to provide all sources installed with the
compiled forms, but some people prefer to keep sources non-installed, just
to spare disk space, say.  What would be ideal is a way for `compile_dir'
to produce `.pyc' in the build hierarchy directly from the source hierarchy
at `make all' time, and that `make install' just move these ready `.pyc'
files into the installation directories.

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





From prestonlanders at my-deja.com  Wed Nov 17 13:07:24 1999
From: prestonlanders at my-deja.com (Preston Landers)
Date: Wed, 17 Nov 1999 18:07:24 GMT
Subject: list.without()?
References: <38319104.F54AECEF@appliedbiometrics.com> <NDBBIAPHEKGOBIHCBLIIEEOOCAAA.mcfletch@vrtelecom.com>
Message-ID: <80uqso$66a$1@nnrp1.deja.com>

In article <NDBBIAPHEKGOBIHCBLIIEEOOCAAA.mcfletch at vrtelecom.com>,
  "Mike Fletcher" <mcfletch at vrtelecom.com> wrote:

> Conclusions:
> 	First is the posted algo (while,try,except,break) <without>
gives very
> good performance for the few-in-many case but very poor performance
for the
> many in many cases for large values of many
> 	Second is a for loop with == test and append <without2> --
gives decent
> (but not spectacular) results whenever memory effects are not a big
factor,
> not really a contender.
> 	Third is the lambda solution (which does surprisingly well)
<without3> --
> gives acceptable but not great performance all over, seems strange
that
> function-call-overhead is so minor a factor.
> 	Fourth is a backwards-index-counting scheme with del
<without4> -- gives
> marginally better performance than third on few-in-many case,
marginally
> worse than third in many-in-many case.
> 	Fifth is Preston's remove function <remove> -- gives best
results for
> few-in-many case, again, very poor results in many-in-many case
>
> Summary:
> 	Looks like the fourth algo is likely the "winner" in this set
(assuming
> few-in-many is the most common case, but that many-in-many is a
potential
> case).  Third might be preferred due to the simplicity of the
implementation
> and similar speed.  For maximum few-in-many speed, the fifth is the
most
> appropriate.

Excellent work, Mike.  Stuff like this is certainly good to know when
working a lot with lists! It looks like one conclusion we can draw from
this is that matching an algorithm to a data set is not always a
straightforward task.  It requires careful thought, familiarity with
the expected cases (few-in-many, many-in-many, etc) and most
importantly, experimentation.  Thanks again!

---Preston

--
|| Preston Landers <prestonlanders at my-deja.com> ||


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



From a.eyre at optichrome.com  Mon Nov  1 12:57:43 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Mon, 1 Nov 1999 17:57:43 -0000
Subject: Looking for example of calling C++ from Python
In-Reply-To: <381ca15e$0$235@nntp1.ba.best.com>
Message-ID: <000d01bf2492$987a72d0$3acbd9c2@peridot.optichrome.com>

> I am new to python and would like to know if there is any example 
> of how to use C++ classes from Python.

Have a look at: http://www.python.org/sigs/c++-sig/

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From thomas at bibsyst.no  Wed Nov  3 02:18:45 1999
From: thomas at bibsyst.no (Thomas Weholt)
Date: Wed, 03 Nov 1999 08:18:45 +0100
Subject: Need Berkley DB!!
References: <381D5E84.98931682@bibsyst.no> <hAQT3.488$bu3.39846291@news.randori.com>
Message-ID: <381FE1D5.C0F721A6@bibsyst.no>

Robin Dunn wrote:

> > After I spent most of my weekend trying to get Berkley DB to work with
> > Python on my Linux ( Suse 6.1) machine, I`ve now given up completly.
> > I`ve tried to install DB.1.85. and DB.2.7.7, neither worked, either
> > crashing under installation ( which of course, is not Python
> > related ) or when I tried to test the installation with the examples in
> > the documentation ( the wrapper for the 2.7.5-module listed at
> > www.python.org, under modules, databases ).
>
> Have you tried my DB module at http://starship.python.net/crew/robind/ ?
> Assuming you can get DB 2.7.x built and installed, this module should do
> what you need.
>

Yes, I`ve tried it and the installation did go smooth, but when I tried to
use the example in the documentation Python barfed some errors at me, then
died with a scream. I can probably recreate the errors and send you a copy
of the output.

I think the problem might be solved if I, sigh, re-install Red Hat 6.1 and
trash that oversized, over-featured Suse 6.2 Linux distro I got now. The Red
Hat version of Python came with all that stuff installed. There`s not enough
hours in a day to keep working on things like this. At least not right now.

Thanks for your interest though.

Best regards,
Thomas Weholt




From mywsid at datawise.net  Sat Nov  6 03:13:04 1999
From: mywsid at datawise.net (mywsid at datawise.net)
Date: Sat, 06 Nov 1999 08:13:04 GMT
Subject: Get Noticed!!
Message-ID: <kqRU3.46477$23.1759836@typ11.nn.bcandid.com>

http:/www.mightybiz.com




From levicook at earthlink.net  Sat Nov 27 17:20:43 1999
From: levicook at earthlink.net (Levi Cook)
Date: Sat, 27 Nov 1999 16:20:43 -0600
Subject: is there a safe/easy way to search and replace text w/in an html page?
Message-ID: <3840593B.2D1A2E30@earthlink.net>

Hello,

I'm new to python and I've been tasked w/ writing a script that fetches
a web-page and highlights the term that the user searched for. Passing
the URL and search term to my script was a cinch and fetching the page
was also quite easy.

The crux of my problem lies with establishing a strategy for searching
out the users term and replacing it with css tags surrounding the
original term.
At first glance the re module seemed sufficient (and maybe still is ),
but it doesn't really have any knowledge of HTML semantics. My next
venture was to look at the htmllib module. This appears to parse the
html fine, but i'm still at a bit of a loss on what this actually does
for me, particularly how do I get my parser to output my page.

Does anyone know where I can find code that does anything remotely like
this?
Does anyone know where I can find good reference on the htmllib module,
the python library reference is pretty bleak in this area.

Thanks in advance for you time and help.
Levi Cook
levicook at earthlink.net


If anyone's interested here's a bit of the code i'm using for this...

<!--$

startTag = '<font class = highlight><strong>'
endTag = '</strong></font>'
cssTag = '<style type="text/css">.highlight {background-color:
#FFFF99}</style>'

if query.has_key("qt"): qt = string.join(query["qt"],',')
else qt = ""

import urllib
if query.has_key("address"): address = query["address"][0]
else address = urllib.quote("http://papajo.com")
address = urllib.unquote(address)

import urlparse
URL2Fetch = urlparse.urlparse(address)

# fetch the page
import httplib
h = httplib.HTTP(URL2Fetch[1])
h.putrequest('GET', URL2Fetch[2])
h.putheader('Accept', 'text/html')
h.putheader('Accept', 'text/plain')
h.endheaders()
errcode, errmsg, headers = h.getreply()
f = h.getfile()
data = f.read()
f.close()

# insert <base> so that all embeded URL's resolve correctly
termstring = ''(^|\\W)(<head>)($|\\W)'
termpat = re.compile(termstring, re.IGNORECASE)
data = re.sub(termpat, '<head>\n<base href="' + address + '">\n' +
cssTag, data)

-->

<!--$write(data)-->





From AmanuelZmelik at news.vex.net  Sun Nov 14 04:02:34 1999
From: AmanuelZmelik at news.vex.net (Amanuel Zmelik)
Date: 14 Nov 1999 09:02:34 GMT
Subject: itoa on python?
References: <382e77d5.26188609@news.remarq.com>
Message-ID: <80ltra$2rb0$1@hub.org>

On Sun, 14 Nov 1999 08:52:03 GMT, Gusdvg <gusdvg at usa.net> wrote:
>How can I change an integer (or any other type of number) to a string?

Here's a few......

>>> str(1234)
'1234'
>>> `1234`
'1234'
>>> "%s" % 1234
'1234'





From prestonlanders at my-deja.com  Thu Nov  4 16:56:33 1999
From: prestonlanders at my-deja.com (Preston Landers)
Date: Thu, 04 Nov 1999 21:56:33 GMT
Subject: return boolean from functions
References: <7vrq9b$hmr$1@nnrp1.deja.com>
Message-ID: <7vsvef$elo$1@nnrp1.deja.com>

Is this not what you want?

def foo(bar):
   if bar == "spam":
      return 1
   else:
      return None

if foo("spam"):
   print "We have spam!"

As you can see, the return value of foo() is evaluated in a true-false
sense by the if statement.  So, return None or 0 when you mean false,
or return 1 or virtually anything else when you mean true.

Alternatively, you can do this:

TRUE=1
FALSE=0

def foo(bar):
  if bar == "spam":
     return TRUE
  else:
     return FALSE

hope this helps,

---Preston

In article <7vrq9b$hmr$1 at nnrp1.deja.com>,
  a_olme at my-deja.com wrote:
> Hello.
>
> Is it possible to return booelan from functions other than return 1==1
> for true or return 1==2 for false?
>
> //Anders Olme
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
>

--
|| Preston Landers <prestonlanders at my-deja.com> ||


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



From tbryan at python.net  Wed Nov  3 21:48:37 1999
From: tbryan at python.net (Thomas A. Bryan)
Date: Wed, 03 Nov 1999 21:48:37 -0500
Subject: Tkinter and Solaris 7 Sparc [Q]
References: <7vqblc$ham$1@nnrp1.deja.com>
Message-ID: <3820F405.4A1C8767@python.net>

exeq at my-deja.com wrote:
> I am trying to compile Python 1.5.2 with tkinter on Solaris 7 (Sparc).

A word of encouragement: I have successfully built Python 1.5.2 on
Solaris 7.
I can't remember whether I used gcc or Sun's compiler.

> have installed TCL/TK 8.0.5, GCC 2.8.1 / Sun Workshop 5 (I tryied with
> both compilers).
> 
> I just get an error saying that libX11.so.6.1 is needed. Is there a way
> to build tkinter without X11? Or I just need to download and install
> X11?

First, I ask the "obvious" question.  Did you edit the Modules/Setup
file 
(or, better, did you make a Setup.local file)?  If so, did you make sure 
to use the entries in the Tkinter portion that are specific to Solaris?

(from Setup)

# *** Uncomment and edit to reflect where your X11 header files are:
#-I/usr/X11R6/include \
# *** Or uncomment this for Solaris:
       -I/usr/openwin/include \
.....
# *** Uncomment and edit to reflect where your X11 libraries are:
#-L/usr/X11R6/lib \
# *** Or uncomment this for Solaris:
       -L/usr/openwin/lib \

If you already did that correctly, is libX11 in that directory?  That is 
use 'ls' and 'find' to look for the library.  Is it there?

---Tom



From robin at alldunn.com  Thu Nov  4 14:35:05 1999
From: robin at alldunn.com (Robin Dunn)
Date: Thu, 4 Nov 1999 11:35:05 -0800
Subject: Extending Python dynamically
References: <3F8BA119B48AD11197470008C7A493CB98B636@HQOBGEX1> <3821D9B0.56C01339@iqsoft.hu>
Message-ID: <melU3.901$bu3.58261916@news.randori.com>

> Thank you for your help. The method you suggested works perfectly!
> However, I do not understand completely how the Python interpreter finds
out
> that it has to load an so file. Normally -- I believed -- when Python sees
an
> "import xxx" statement it looks for a xxx.py, xxx.pyc or xxx.pyo file in
the
> $PYTHONPATH directories or an xxx subdirectory having an __init__.py file,
but
> not an xxx.so file. This means that if I add the current directory to the
> $PYTHONPATH (such as export PYTHONPATH='.' Python will find any arbitrary
so
> file in the $PWD?

It will also load xxxmodule.so as well, but .so files only get imported if
they have an initxxx() function.


--
Robin Dunn
Software Craftsman
robin at AllDunn.com
http://AllDunn.com/robin/
http://AllDunn.com/wxPython/  Check it out!







From kvijayan at calum.csclub.uwaterloo.ca  Sat Nov 13 23:13:08 1999
From: kvijayan at calum.csclub.uwaterloo.ca (Kannan Vijayan)
Date: 14 Nov 1999 04:13:08 GMT
Subject: GNU readline module
Message-ID: <80lcsk$99g$1@watserv3.uwaterloo.ca>

hi

I'm trying to get the readline module
to work in python.
I have an app to which I want to add
tab-completion of commands and other things.

The docs for the GNU readline module are
extremely sparse.  Most of what I know about
the module are derived from looking at it's
source, looking at how "rlcompleter" uses it,
and my own sparse knowledge of the C readline
implementation itself.


As far as I can see, there is no hook to the
"readline(char *PROMPT)" function in the
readline module.  Am I missing something?

I have, however, gotten my completer function
to work well from within the interactive
interpreter environment.  What I'm having
trouble with is getting it to work from a
non-interactively interpreted application in
it's own happy .py file.  I assume there would
be some function which I could call to read a
line, which in turn would call my completer
function to do the job when a tab is pressed.

I have a nagging notion that maybe the readline
module is only for use from withing the
interactive interpreter.  IS this true? if so..
that's extremely dissapointing.

thank you
-kannan



From raptor_99_uk at my-deja.com  Wed Nov 10 10:36:36 1999
From: raptor_99_uk at my-deja.com (raptor_99_uk at my-deja.com)
Date: Wed, 10 Nov 1999 15:36:36 GMT
Subject: Namespace question...
References: <7vqgib$l8v$1@nnrp1.deja.com>
Message-ID: <80c3e4$vpq$1@nnrp1.deja.com>

In article <7vqgib$l8v$1 at nnrp1.deja.com>,
  oofoe at my-deja.com wrote:
> Hi!
>
> How can you access the global namespace in
python 1.5.1? I'd like to get
> a version string which is declared in the main
script file. However,
> each module also has a version string which
overrides the original one.
> I tried to access it as __main__.Version,
similarly to $main::Version in
> Perl, which is what I'm more familiar with. Of
course, it didn't work.
>
> I have something like this:
>
> module file foo.py:
> ------------------------------
> Version = '$Id$'
>
> def bar():
>    print "Version is: ", __main__.Version
>
> main program:
> ------------------------------
> Version = '$Id$'
>
> import foo
>
> foo.bar()
>
> -------------------------------
>
> I'd certainly appreciate any pointers. The
namespace stuff seems to get
> a little hazy around issues like this.
>
> Thanks!
>
> Jos'h
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
>

just declare Version to be global inside bar() as
follows:

def bar():
    global Version
    print "Version is: ", Version


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



From ionel at psy.uva.nl  Mon Nov  8 14:45:48 1999
From: ionel at psy.uva.nl (Ionel Simionescu)
Date: Mon, 8 Nov 1999 20:45:48 +0100
Subject: Fastest GUI toolkit?
References: <01bf29f1$cbfd5e60$5a41c3d1@Cyberus.cyberus.ca>
Message-ID: <80796r$5ml@mail.psy.uva.nl>

Gordon Williams <g_will at cyberus.ca> wrote in message
news:01bf29f1$cbfd5e60$5a41c3d1 at Cyberus.cyberus.ca...
|
| I have an application where I need to draw many lines fast and repeatedly
| (an oscilloscope).  The number of lines ranges from about 1000 to 16000.
| Currently I have the application using Tkinter but I am finding it too
| slow.  I am currently using W95 as an os, but would prefer to make the
code
| portable across other os if possible.  Right now speed is more important
| than portability so I would be willing to give this up.
|
| I would like to draw 1000 lines in 1/10 of a second on a P166 machine.
(Am
| I way off the mark??)  I was hoping that someone would have tried
different
| packages and could give me pointers in the right direction.


Hi Gordon,

I've just tried your test with a simple wxPython application and it almost
makes it.

My system is Windows NT, P200, 64M.

The test app draws 1400 lines in .25 seconds (with the same pen).
For 1000 lines, it takes 0.18 seconds.

wxPython is also offering an easy to use OpenGL interface, from which I
expect much better performance.

Given your needs, I suggest that you should consider seriously using OpenGL,
rather than trying your own optimizations. This choice would give you good
and scalable performance irrespective of the platform.

Just naive GDI calls won't do, I'm afraid.
However, if you need something quick, you may consider to draw in a memory
DC (i.e. a bitmap) first and copy the end result to the screen. That will
hopefully make the time drop under your .1 second mark.

---
ionel

References:
wxPython:
http://alldunn.com/wxPython/







From jimbag at kw.igs.net  Mon Nov 15 23:24:19 1999
From: jimbag at kw.igs.net (JB)
Date: Tue, 16 Nov 1999 04:24:19 GMT
Subject: ASCII delimited files
References: <942231502.1914033961@news.interbusiness.it>
Message-ID: <3830DC72.1852529A@kw.igs.net>

Roger Irwin wrote:
> 
> Is there any function or module available for parsing ASCII delimited files,
> before I go and re-invent the wheel writing my own.
> 
> BTW, I know it is simple to do, but I was thinking along the lines that a
> 'standard' could easily be substituted for a C version for large files.

I wrote something that may be of interest. It does CSV, Fixed and
delimited ASCII files (amoung other things). It's called dataSlayer (i
know it's a bit dramatic)

http://www.kw.igs.net/~jimbag

cheer
jim

-- 
-----------------------------------------------------------
Machine-Independent, adj.:
	Does not run on any existing machine.



From pinard at iro.umontreal.ca  Tue Nov 30 15:49:28 1999
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 30 Nov 1999 15:49:28 -0500
Subject: Python Make
In-Reply-To: Helge Hess's message of "Tue, 30 Nov 1999 19:29:04 +0100"
References: <3843E558.8AC91A97@strs.co.uk> <38441770.80C33C93@mdlink.de>
Message-ID: <oq7lizaenb.fsf@titan.progiciels-bpi.ca>

Helge Hess <helge.hess at mdlink.de> writes:

> Ian Clarke wrote:

> > Has anyone considered using Python as a replacement for the "make" [...]

> YES! I had the same idea :-) [...] The negative aspect is that one
> needs python for compilation then (eg you can't compile python using
> 'python' makefiles).

Detail!  Given you have some `dryrun' option, just use it for generating
a distributable batch script, that could then be used for bootstrap.

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





From ScherBi at BAM.com  Tue Nov 16 16:01:01 1999
From: ScherBi at BAM.com (ScherBi at BAM.com)
Date: Tue, 16 Nov 1999 16:01:01 -0500
Subject: problems with __getattr__ & __setattr__
Message-ID: <3F8BA119B48AD11197470008C7A493CB98B67E@HQOBGEX1>

Michael Hudson wrote:

>      def __getattr__(self, key):
>          try:
>              return self.__dict__["eggs"][key]
>          except:
>              raise
> 
> is probably safer. What are the try/catch for? They aren't doing
> anything here!

Thanks.  I came upon your this solution in the interval.
The try/except were not relevent in the sample, but are in my real code.
sorry. 8-)

> __getattr__-and-__setattr__-are-so-much-fun-we-don't-mind-the-
> performance-hit-ly y'rs - Michael

Yes, I had to convince a colleague (it wasn't hard) that the syntactic sugar
they produce are worth the __dict__ mangling ugliness. :)

Thanks again.

Bill



> -----Original Message-----
> From: Michael Hudson [mailto:mwh21 at cam.ac.uk]
> Sent: Tuesday, November 16, 1999 3:41 PM
> To: python-list at python.org
> Subject: Re: problems with __getattr__ & __setattr__
> 
> 
> ScherBi at BAM.com writes:
> 
> > I am having trouble using __getattr__ and __setattr__ in 
> some classes.  I've
> > boiled it down to the code below.
> > It dumps core under 1.5.2 on Linux or WinNT.
> > 
> > I think maybe it's looping, if so, how does one go about 
> doing this? 
> > (I'm assuming it's clear enough what I'm trying to do.)
> > 
> > If I comment out the __gettattr__ operation, it doesn't 
> dump core.  Instead
> > I get an AttributeError: eggs raised at the 'self.eggs[key] 
> = value' call in
> > __setattr__.  This is what leads me to belive there's 
> something circular
> > going on here.
> > 
> > 
> > Thanks,
> >  Bill
> > 
> > -------------------------------------------------------------
> > 
> > class spam:
> > 
> >     def __init__(self):
> >         self.eggs = {}
> 
> Hint: what does this call?
>      
> >     def __getattr__(self, key):
> >         try:
> >             return self.eggs[key]
> >         except:
> >             raise
> >         
> >     def __setattr__(self, key, value):
> 
> It calls this!
> 
> >         try:
> >             self.eggs[key] = value
> 
> And this calls __getattr__ which looks for self.eggs, which calls
> __getattr__, which looks for self.eggs, which calls __getattr__, which
> looks for self.eggs, which calls __getattr__, which looks for
> self.eggs, which calls __getattr__, which looks for self.eggs, which
> calls __getattr__, 
> Segmentation fault (core dumped)
> 
> >             return 0
> >         except:
> >             raise
> > 
> > 
> > 
> > if __name__ == '__main__':
> > 
> >     s = spam()
> >     print 'instance of spam created.' 
> > 
> 
> try
> 
>     def __init__(self):
>         self.__dict__['eggs'] = {}
> 
> instead; also 
> 
>      def __getattr__(self, key):
>          try:
>              return self.__dict__["eggs"][key]
>          except:
>              raise
> 
> is probably safer. What are the try/catch for? They aren't doing
> anything here!
> 
> __getattr__-and-__setattr__-are-so-much-fun-we-don't-mind-the-
> performance-hit-ly y'rs - Michael
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 



From aahz at netcom.com  Thu Nov 25 21:51:18 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 26 Nov 1999 02:51:18 GMT
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <al8yabo9fm3.fsf@sirppi.helsinki.fi> <86so1v4jlj.fsf@g.local> <al8hfiaan8p.fsf@sirppi.helsinki.fi>
Message-ID: <81ksj6$lif$1@nntp9.atl.mindspring.net>

In article <al8hfiaan8p.fsf at sirppi.helsinki.fi>,
Markus Stenberg  <mstenber at cc.Helsinki.FI> wrote:
>
>From trivial example - do you consider (+ 2 (* 3 4)) more readable than 2 +
>3 * 4 most languages (including Dylan) allow me to do?

I think that should be written as "3 4 * 2 +".  ;-)
--
                      --- Aahz (@netcom.com)

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

Question authority: ask me anything!



From smst at quantisci.co.uk  Tue Nov 16 04:57:57 1999
From: smst at quantisci.co.uk (Steve M S Tregidgo)
Date: Tue, 16 Nov 1999 09:57:57 GMT
Subject: Memory not being released?
Message-ID: <199911160957.JAA12725@cobweb.quantisci.co.uk>

Thanks.  We already have a persistent object system, but I'll take a
look at shelve anyway.

Darrell wrote:
> Here's an idiom I like in this case.
> Store your objects in a dict based on id(obj), just as you've done.
> Now if objects must reference each other use the id(obj)
> and look up the obj. And delete them from the dict when needed.
> Yes it's slow but only use it in these big ugly cyclic data
structures.
> 
> But notice how fast memory gets used up.
> This eats 84meg.
> >>> dict={}
> >>> for x in xrange(1000000):
> ....     dict[x]=(x,x+1)
> ....
> >>>
> 
> You might take a look at shelve. It gives you a persistent
dictionary.
> I've been playing with a flat file manager that can reclaim freed
space.
> 
> --
> --Darrell
> Steve Tregidgo <smst at bigfoot.com> wrote in message
> news:80pdvo$99r$1 at nnrp1.deja.com...
> > Hi there,
> >
> > I'm having problems with a large process, and I'd really
appreciate
> > some help in tracking down the cause of its bloatedness.
> >
> 
> 
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list



From jsight at mindspring.com  Sun Nov 28 20:56:19 1999
From: jsight at mindspring.com (Jesse D. Sightler)
Date: Sun, 28 Nov 1999 20:56:19 -0500
Subject: http://www.python.org/ blank ?
References: <HJYc$DA+GcQ4EwYP@jessikat.demon.co.uk>
Message-ID: <3841DD43.C0C07A03@mindspring.com>

Close all Netscape windows.  Reopen Netscape.  Go to
http://www.python.org/.  Magically it will appear.  :-)

This has been a weird little display glitch in many versions of Netscape
for years where every once and a while a page will load completely...
and then NOT display at all.

Robin Becker wrote:
> 
> for some reason I see http://www.python.org/ as a blank page. Any ideas?
> --
> Robin Becker

-- 
---------------
Jesse D. Sightler
http://www.biddin.com/



From helka.folch at edf.fr  Tue Nov  9 12:56:37 1999
From: helka.folch at edf.fr (Helka FOLCH)
Date: Tue, 09 Nov 1999 18:56:37 +0100
Subject: Max Size of Array when using Numeric.py
Message-ID: <38286054.C8612EB5@edf.fr>

I've tried to create an array object whose shape is (5500,200) but I get
a memory error. Is there a maximum size for array objects?

Maybe there's a better way of creating the array than what I did :
My program reads the array which is in another format from a file. It
transforms the array to the right format and holds it in a variable(mv)
as a string. Then I just do:

m = array(eval(mv),Float)

Thanks a lot,
Helka




From matz at netlab.co.jp  Thu Nov  4 00:22:08 1999
From: matz at netlab.co.jp (Yukihiro Matsumoto)
Date: 04 Nov 1999 14:22:08 +0900
Subject: A Comparison of Python and Ruby Extension Interfaces (Re: Comparison between Python and "Ruby")
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp>
 <7vjtrc$hmm$1@newshost.accu.uu.nl> <38204614.2EAC6051@compaq.com>
 <021119992327527678%pereira@research.att.com>
 <38215CE7.70C3D560@compaq.com>
 <031119992347056096%pereira@research.att.com>
Message-ID: <87k8nyvnjz.fsf@ev.netlab.co.jp>

Fernando Pereira <pereira at research.att.com> writes:

|In my experience, there's a big difference between what you and I call
|"sensible" and what compiler and OS developers try to get away with,
|not to mention the vagaries of particular hardware. For example, what
|is the state of the stack, registers, etc. when a Unix signal handler
|is called? What if that handler (maybe by calling back into the
|higher-level language) forces a GC?

I implemented Ruby to avoid GC (and longjmp) during handlers.
See code for detail, if you are interested.  We haven't met such
problems, yet.



From wtanksle at hawking.armored.net  Mon Nov  1 11:43:58 1999
From: wtanksle at hawking.armored.net (William Tanksley)
Date: 1 Nov 1999 16:43:58 GMT
Subject: Improved struct module
References: <000001bf15fc$3397ee00$ba2d153f@tim> <0E16861EE7BCD111BE9400805FE6841F0BEB2650@c1s5x001.cor.srvfarm.origin-it.com> <oqso2phk2j.fsf@titan.progiciels-bpi.ca>
Message-ID: <slrn81s9mj.d51.wtanksle@hawking.armored.net>

On 01 Nov 1999 18:24:20 -0500, Fran?ois Pinard wrote:
>"Robin Boerdijk" <robin.boerdijk at nl.origin-it.com> writes:

>> I must admit that I am also one of those lazy bastards that uses his
>> newsreader's "find any messages by Tim Peters"-function to locate
>> interesting threads.

>So do I, but even more lazily!  Gnus bolds Tim's messages automatically
>for me, when preparing the threaded summary for the Python mailgroup. :-)

Ah, but does your newsreader automatically delete all the other messages?
Hah!  Some time-saver it is.

I used to use nn (NoNews).  Now THAT was a good newsreader -- or news
ignorer as the case may be.  In most newsreaders the default is to read
every message; in NN the default is to ignore every message.  A pity I
couldn't set it up to forge cancels for every message and save everyone
else's time ;-).

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

-- 
-William "Billy" Tanksley



From python-list at teleo.net  Wed Nov 24 18:39:10 1999
From: python-list at teleo.net (Patrick Phalen)
Date: Wed, 24 Nov 1999 15:39:10 -0800
Subject: newbie question
In-Reply-To: <943469319.838266356@news.teleweb.at>
References: <943469319.838266356@news.teleweb.at>
Message-ID: <99112415413404.01208@quadra.teleo.net>

[Markus Krutz, on Wed, 24 Nov 1999]
:: How the f*** do I create a string containing a *single* "\" ?

Well, you could always try escaping it: "\\".

You might also try reading the lovely tutorial Guido worked so hard to
prepare: http://www.python.org/doc/tut/tut.html



From quinn at pfennig.ugcs.caltech.edu  Wed Nov 17 20:07:15 1999
From: quinn at pfennig.ugcs.caltech.edu (Quinn Dunkan)
Date: 18 Nov 1999 01:07:15 GMT
Subject: help redefining print statement
References: <80v32d$cmg$1@nnrp1.deja.com>
Message-ID: <slrn836kbk.2f9.quinn@pfennig.ugcs.caltech.edu>

On Wed, 17 Nov 1999 20:26:59 GMT, curtin at my-deja.com <curtin at my-deja.com> wrote:
>hi,
>i'd like to redefine the "print" statement in a module. essentially
>i'm looking to do a  "#define like" replacement on a module level.
>
>i'd add at the top of the module:
>
>if debug:
>    TestingPrint=print
>else:
>    TestingPrint=eval("#") or TestingPrint="#" or ...
>
>TestingPrint "blah, blah %d, %s" % (i, s)
>
>yes, i'd like to not make the function call overhead in the non-debug
>case.
>
>i've tried a few different ways to incorporate this functionality,
>and scrounged around deja without any success.

You can't redefine print, it's a reserved work built into the language.  My
usual tactic is to use my dprint() function which checks for the debug
variable, and I don't worry about function call overhead.  But if you've got
one it in a tight loop or something, you could do:

if __debug__:
    print 'debugging info'

When python is run with -O, it won't compile things under if __debug__ (see
ref manual 6.2).

maybe-tim-could-write-a-function-that-comments-out-its-caller



From stidolph at origin.ea.com  Mon Nov 22 10:12:03 1999
From: stidolph at origin.ea.com (Stidolph, David)
Date: Mon, 22 Nov 1999 09:12:03 -0600
Subject: Getting PythonWin debugger working with an embedded app.
Message-ID: <11A17AA2B9EAD111BCEA00A0C9B41793034AAF00@molach.origin.ea.com>

I am trying to get the Python Win debugger working with our app.  We have
embedded Python and using it as a script language.  We are using the debug
version of the Python DLL for Windows.

I copied the pywin directory to our python system directory (plus adding
that to the search path) and put the following code in one of our scripts
(to be activated from a menu item within our ui):

import pywin.debugger
try:
    a = 1 / 0
except:
    pywin.debugger.post_mortem()

I get the following output:

Traceback (innermost last):
  File "python\client\Interface\UIMenu.py", line 272, in DoMenuCommand
    command(x,y,flags)
  File "python\client\Interface\UO2GameWindow.py", line 304, in OnMenuDebug
    pywin.debugger.post_mortem()
  File "C:\Program Files\Python\Pythonwin\pywin\debugger\__init__.py", line
70, in post_mortem
    p = _GetCurrentDebugger()
  File "C:\Program Files\Python\Pythonwin\pywin\debugger\__init__.py", line
9, in _GetCurrentDebugger
    import debugger
  File "C:\Program Files\Python\Pythonwin\pywin\debugger\debugger.py", line
19, in ?
    import win32ui
ImportError: No module named win32ui

I tried putting the contents of the PythonWin directory in the pywin
directory and it still has the same message.
In looking at the message I see that it is going to the installed place of
Python and PythonWin instead of my search path.  I can see the win32ui.py
module, and it is directly in the Python search path, so what am I doing
wrong?

Any advice or suggestions would be truly welcome.  Getting the debugger
working would be a huge step forward for us.


Thanks for you help,

David.





From gerrit.holl at pobox.com  Tue Nov 16 11:07:28 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Tue, 16 Nov 1999 17:07:28 +0100
Subject: SMTP
In-Reply-To: <80rn4o$bnl$1@news.fsu.edu>
References: <80rn4o$bnl$1@news.fsu.edu>
Message-ID: <19991116170728.A30970@optiplex.palga.uucp>

Glenn Kidd wrote:
> Does anyone know where some good info on Python's smtplib?

Well, er... what about the default module index?

> I have looked at
> http://www.python.org but I was wondering if there was any more info out
> there.  Any help would be appreciated.

Isn't it enough?

body = '''From: me <friend at somewhere.org>
To: my brother <brother at somehost.com>
Subject: hello!
X-Mailer: a python script

This is the bodddddddddyyyyyyyyyy........
'''

regards,
Gerrit.

-- 
We are using Linux daily to UP our productivity - so UP yours!
(Adapted from Pat Paulsen by Joe Sloan)



From tbryan at python.net  Sun Nov  7 00:49:24 1999
From: tbryan at python.net (Thomas A. Bryan)
Date: Sun, 07 Nov 1999 00:49:24 -0500
Subject: making the Python case to my dept.
References: <9911051740130D.02530@william.jinsa.org> <3823FC43.CF7FBD71@python.net> <m3puxno6g0.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <382512E4.21A48861@python.net>

Michael Hudson wrote:

> ``import rlcompleter''!!

Thanks.  I used to use this, but I took a break from Python, changed 
environments, and forgot about it.  

> [snippo two]
> > Do math majors or engineers learn their critical job skill's in their
> > first classes?
> 
> What are mathematicians' critical job skills? I'm curious, being one,
> and not having being taught anything resembling a `critical job skill'
> yet, and happy about it.

It depends what job you're looking for. :)

I was thinking more about applied math and engineering students, 
but if you're intending to stay in academia, then being able to 
write and explain proofs is a job skill.  So is teaching ability.

I happen to have a math degree, and I specialized in topology with 
some emphasis on knot theory.  When I was originally looking for jobs, 
I emphasized my problem-solving skills and my ability to deal with 
large doses of abstraction and technical literature.

I've found that experience hammering down details to make an air-tight, 
pedantic proof really helps when I'm trying to debug a program.

It-sounds-better-than-"I-can-tie-my-shoes"-ly yours

---Tom



From fdrake at acm.org  Fri Nov  5 11:53:07 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Fri, 5 Nov 1999 11:53:07 -0500 (EST)
Subject: 64 bit ints for SNMP counter64 variables
In-Reply-To: <7vq8o4$18j$1@news.tamu.edu>
References: <7vq8eg$4k$1@news.tamu.edu>
	<7vq8o4$18j$1@news.tamu.edu>
Message-ID: <14371.2931.226939.602387@weyr.cnri.reston.va.us>

Eric Nelson writes:
 > oops... posted too quick... found long() just after i hit send... sorry

Eric,
  There's also the mpz module if you can use the GNU arbitrary-
precision math library (I don't know how portable that is, but I
imagine it should be fairly portable).


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From R.Brodie at rl.ac.uk  Tue Nov  9 08:20:46 1999
From: R.Brodie at rl.ac.uk (Richard Brodie)
Date: Tue, 9 Nov 1999 13:20:46 -0000
Subject: Graphics.
References: <3826E949.BBA1BAFD@imf.au.dk> <091119991052098844%kantel@mpiwg-berlin.mpg.de>
Message-ID: <80973f$poq@newton.cc.rl.ac.uk>

J?rg Kantel <kantel at mpiwg-berlin.mpg.de> wrote in message

> If you want to use Tkinter and maybe you can read German than I've a
> nice example at my website at
>
> <http://www.mpiwg-berlin.mpg.de/staff/kantel/pythonmania/chapter04.html>
>

...and as everyone knows, if you haven't done Python in German, you
haven't done Python :)

Ich bin ein Holzf?ller und f?hl mich stark,
Ich schlaf des Nachts und hack am Tag.





From gerrit.holl at pobox.com  Mon Nov  1 15:47:46 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Mon, 1 Nov 1999 21:47:46 +0100
Subject: Bug in standard cursesmodule?
Message-ID: <19991101214746.A30582@optiplex.palga.uucp>

Hello,

A does not work,
B does work!

--- *** A *** ---
#!/usr/bin/python
import curses

stdscr = curses.initscr()
stdscr.addch(24, 79, ord('X'))
stdscr.refresh()
curses.endwin()

--- *** A *** ---
--- *** B *** ---
#include <ncurses.h>

main () {
        initscr();
        mvaddch(24, 79, 'X');
        refresh();
        endwin(); }
--- *** B *** ---

I think it's a bug... Am I right?

When I run A, it says:
error: [mv]waddstr() returned ERR

regards,
Gerrit.



From fdrake at acm.org  Mon Nov 15 10:26:28 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Mon, 15 Nov 1999 10:26:28 -0500 (EST)
Subject: Random text generator
In-Reply-To: <01e101bf2dcd$f4bb6290$f29b12c2@secret.pythonware.com>
References: <199911120349.OAA01177@envy.fulcrum.com.au>
	<y0jk8nntzvn.fsf@vier.idi.ntnu.no>
	<01e101bf2dcd$f4bb6290$f29b12c2@secret.pythonware.com>
Message-ID: <14384.9764.78515.201565@weyr.cnri.reston.va.us>

Fredrik Lundh writes:
 > it would probably be a bit clearer (at least to newcomers)
 > if "random" is was the official interface, and whrandom was
 > treated as an implementation module.  but it doesn't look
 > that way in the current version of the library reference...

  This is already on my list to fix before the next documentation
release.

 > (and the eff-bot guide is obviously wrong on this one. not
 > sure I'll add that to the errata, though ;-)

  Add it so people will get the message!


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From fdrake at acm.org  Fri Nov 19 09:12:01 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Fri, 19 Nov 1999 09:12:01 -0500 (EST)
Subject: Python Chinese
In-Reply-To: <19991119080202.A439@better.net>
References: <19991119121544.5006.qmail@hotmail.com>
Message-ID: <14389.23217.980958.63607@weyr.cnri.reston.va.us>

On Fri, Nov 19, 1999 at 01:15:44PM +0100, Ulf Engstr?m wrote:
 > Is there any way to use Python with Chinese and other asian chars?
 > If not, are there any plans to make this compability?

Ulf,
  While Python is 8-bit clean, there is not support for Unicode or
other wide character sets.  Some extensions are available from various 
members of the community to provide Unicode support, but none are
integrated into the core.
  The next major Python release will include Unicode support and
support for a variety of common encodings.  Support for Asian
encodings will probably be packaged separately.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From mwh21 at cam.ac.uk  Tue Nov 30 06:48:38 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 30 Nov 1999 11:48:38 +0000
Subject: Installation Problem (1.5.2)
References: <3843B816.F4A31F07@attglobal.net>
Message-ID: <m366ykryi1.fsf@atrus.jesus.cam.ac.uk>

Jon McLin <jon_mclin at attglobal.net> writes:

> On my Mandrake 6.0 box, I just upgraded from 1.5.1 to 1.5.2 with a
> fresh compile.  Since the 1.5.1 installation placed the binaries in
> /usr/bin, I used the configure option "--exec_prefix=/usr"

1) you mean --exec-prefix.

2) you may have binaries in /usr/bin, but you also have libraries in
   /usr/lib, not /usr/local/lib. pass --prefix=/usr to configure.

Regards,
Michael



From guido at cnri.reston.va.us  Wed Nov  3 18:33:30 1999
From: guido at cnri.reston.va.us (Guido van Rossum)
Date: 03 Nov 1999 18:33:30 -0500
Subject: A Comparison of Python and Ruby Extension Interfaces (Re: Comparison   between Python and "Ruby")
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp> <7vjtrc$hmm$1@newshost.accu.uu.nl> <38204614.2EAC6051@compaq.com>
Message-ID: <5ld7trtak5.fsf@eric.cnri.reston.va.us>

Greg Ewing <greg.ewing at compaq.com> writes:

> * The names of the Ruby interface routines are often
> shorter than the equivalent Python ones.

Python used to have short names too, and in fact the current naming
scheme was intruduced with great reluctance.  Unfortunately the
realities of embedding Python in other apps required the current
longer names, to avoid name clashes between Python macros, types and
externals and ditto defined by the embedding app.  Extension modules
wrapping large existing libraries also ran into this.

> * Ruby seems to be able to raise exceptions without
> having to explicitly propagate them back out of routines,
> which eliminates a lot of result testing and returning of
> -1's and None's that goes on in Python extension code.

Does it use longjmp to do this?  Or is it written in C++ and using C++
exceptions?  Note that longjmp is a frequent cause of nasty
portability problems (e.g. recently a bug reported on FreeBSD when
combining threads and readline was discovered to be caused by a bug in
the longjmp implementation).

Or perhaps it has macros containing return statements?  (Yuck!)

> Some other observations, in no particular order:
> 
> * In Ruby, routines with a fixed number of arguments can
> be defined so that they are received as separate C
> arguments instead of a vector, e.g.
> 
> static VALUE
> fdbm_store(obj, keystr, valstr)
>     VALUE obj, keystr, valstr;
> 
> This doesn't actually save much code, however, since you
> still have to test and convert each argument.

How does Ruby know how many arguments the C code has???

> * Ruby's method of representing an extension type is a
> whisker more complicated. Instead of allocating a single
> object, you have to create an object of type DATA, which
> contains a pointer to a class and a pointer to a C
> structure.
> 
> * The mechanism for associating methods with object types
> is somewhat simpler in Ruby. In Python, some methods live
> in slots in the type object, while others live in a method
> table or are somehow otherwise accessed through the getattr
> method. In Ruby, all methods, including the ones for built
> in operations, are attached to the object's class, which seems
> to play the role of both the Python type object and method
> table.

I agree that this would be a desirable design for a future Python
implementation.

--Guido van Rossum (home page: http://www.python.org/~guido/)



From jlle at redestb.es  Sun Nov  7 13:48:25 1999
From: jlle at redestb.es (Juan)
Date: Sun, 7 Nov 1999 19:48:25 +0100
Subject: Printing from Python to Windows Printers
References: <fIHU3.641$f66.886@198.235.216.4>
Message-ID: <804hb0$4cb6@SGI3651ef0.iddeo.es>

You can use the class wxPrinter and wxPrintout instaling previously
wxPython.

Un saludo.
Juan Huertas

Robert Parker <rparker at bserv.com> escribi? en el mensaje de noticias
fIHU3.641$f66.886 at 198.235.216.4...
> I am trying to find out how I can print to a windows printer from a Python
> application.  Has anyone used any of the Com modules to gain access to the
> printer api in windows?  Are there any modules that have been written that
> would help save time when we want to send font information only to the
> printer driver not have to write in the escape sequences and print
directly
> to the parallel port?
>
>
>
>





From conversy at lri.fr  Mon Nov 15 10:43:26 1999
From: conversy at lri.fr (Stephane Conversy)
Date: Mon, 15 Nov 1999 16:43:26 +0100
Subject: Tkinter.event.state
Message-ID: <38302A1D.7224CD04@lri.fr>

Hello.

I know that the 'state' filed of a Tkinter event changed whenever I
press
the shift key, or the ctrl key.

I can see and figure out what to test if I want to query if the 'shift'
key
is pressed, but I was wondering whether it exists some constant
to do so. I looked in the doc and in the sources and I couldn't find it.

any tips ?

--
St?phane Conversy
http://www-ihm.lri.fr/~conversy/
mailto:conversy at lri.fr






From matz at netlab.co.jp  Wed Nov  3 21:30:13 1999
From: matz at netlab.co.jp (Yukihiro Matsumoto)
Date: 04 Nov 1999 11:30:13 +0900
Subject: A Comparison of Python and Ruby Extension Interfaces (Re: Comparison   between Python and "Ruby")
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp>
 <7vjtrc$hmm$1@newshost.accu.uu.nl> <38204614.2EAC6051@compaq.com>
 <5ld7trtak5.fsf@eric.cnri.reston.va.us>
Message-ID: <87puxrugy2.fsf@ev.netlab.co.jp>

Guido van Rossum <guido at cnri.reston.va.us> writes:

|Does it use longjmp to do this?  Or is it written in C++ and using C++
|exceptions?  Note that longjmp is a frequent cause of nasty
|portability problems (e.g. recently a bug reported on FreeBSD when
|combining threads and readline was discovered to be caused by a bug in
|the longjmp implementation).
|
|Or perhaps it has macros containing return statements?  (Yuck!)

It's using longjmp.  We haven't met longjmp's nasty portability
problems yet, for 6 years.

|> This doesn't actually save much code, however, since you
|> still have to test and convert each argument.
|
|How does Ruby know how many arguments the C code has???

It's part of the method attribute.  C functions are registered with
the number of arguments they accept.

|> * The mechanism for associating methods with object types
|> is somewhat simpler in Ruby. In Python, some methods live
|> in slots in the type object, while others live in a method
|> table or are somehow otherwise accessed through the getattr
|> method. In Ruby, all methods, including the ones for built
|> in operations, are attached to the object's class, which seems
|> to play the role of both the Python type object and method
|> table.
|
|I agree that this would be a desirable design for a future Python
|implementation.

Surely.
							matz.



From nhv at cape.com  Tue Nov 16 14:12:22 1999
From: nhv at cape.com (Norman Vine)
Date: Tue, 16 Nov 1999 14:12:22 -0500
Subject: FW: [3D Modeller] Final decisions... scripting ?
References: <80rb9c$m87$1@nnrp1.deja.com>
Message-ID: <80sabs$2ld0$1@newsie2.cent.net>

FWD: from 3dmodeller at topica.com

>-----Original Message-----
>From: Joel Utting
>Sent: Tuesday, November 16, 1999 1:21 PM
>To: 3dmodeller at topica.com
>Subject: [3D Modeller] Final decisions... scripting ?
>
>
>To try to wrap up this debate; I don't think we should rely on a
>scripting language, rather use it for extra commands/plugins. Chris'
>mail about not wanting to rely on a scripting layer is correct in my
>opinion.
>
>The Undo feature is easily implemented without a scripting
>layer. Why is
>it necessary ? I can see how it would be useful (later) but I think
>passing a Number relating to the button press/etc (eventually
>the number
>of the script to be called) through the system is faster and simpler to
>begin with - if the number is a 1:1 relation to a function call...call
>the function. If not then call the script interpreter that the number
>relates to with the script that the number refers to and let the script
>do all the more fancy work. Its not inventing a new scripting language
>as scripts with 1:1 relationship are function calls (all our base
>functionality) and anything else calls a script written in a common
>scripting language.
>
>Setting up a script language is what should come later on -
>when we make
>the app more customizeable/flexible.
>
>Perhaps a more detailed plan of how the script interface/editor core
>should work would help - we need this to start coding soon.
>
>Bye - Joel.
>
>
>
>3D Modeller Mailing list ;
>to unsubscribe 3dmodeller-unsubscribe at topica.com
>archives  http://www.topica.com/lists/3dmodeller
>URL ; http://www.woodsoup.org/~sbaker/modellers
>






From cgw at fnal.gov  Tue Nov 16 08:24:16 1999
From: cgw at fnal.gov (Charles G Waldman)
Date: Tue, 16 Nov 1999 14:24:16 +0100 (MET)
Subject: Directory chooser in Tk
In-Reply-To: <19991116071917.A711753@vislab.epa.gov>
References: <38304E78.F92C8596@callware.com>
	<19991116071917.A711753@vislab.epa.gov>
Message-ID: <14385.23296.680071.39959@gargle.gargle.HOWL>

 > Ivan Van Laningham:
 >  |I need a dialog similar to the askopenfilename() dialog, but for
 >  |directories.  Is there such a beast?
 > 

A long time ago I needed this and wound up using Tix.  But I imagine
that by now Tix is deprecated, unsupported, obsolete, dead, etc...

How-about-a-filechooser-that-lets-you-select-multiple-files-ly yr's,
							       
							       cgw




From infotechsys at pivot.net  Tue Nov 30 01:10:32 1999
From: infotechsys at pivot.net (Wayne)
Date: Tue, 30 Nov 1999 01:10:32 -0500
Subject: Help finding Matt Conyway's tutorial
Message-ID: <38436A58.98FEA03@pivot.net>

Hi,
I'm trying to find an address for Matt Conway. I understand he has
a tutorial called "A Tkinter Life Preserver" that I would like to
down load. If there is any other tutorial out there that someone
can point me to I be happy for the info.
TIA.
Wayne




From mcfletch at vrtelecom.com  Mon Nov 29 07:43:20 1999
From: mcfletch at vrtelecom.com (Mike Fletcher)
Date: Mon, 29 Nov 1999 07:43:20 -0500
Subject: [wxPython] Mandelbrot - further improvements
In-Reply-To: <NDBBIAPHEKGOBIHCBLIIKEANCBAA.mcfletch@vrtelecom.com>
Message-ID: <NDBBIAPHEKGOBIHCBLIICEAOCBAA.mcfletch@vrtelecom.com>

Summary:
	Mandelbrot Module, calculates (and displays) Mandelbrot set using Numeric
extensions.  Has been optimised slightly on the wxPython list, module
attached...

Another few 1/10ths of a second improvement (~3.3s down to ~2.9s), as my
part has ceased to have any relevance to wxPython, I should probably switch
to the main list :) , this will be my last post to the wxPython list on this
topic :) .  I added code to the module so that it will run without wxPython
installed (it will use wxPython if it's available, otherwise just prints
status to command line).

Enjoy,
Mike




From S.I.Reynolds at cs.bham.ac.uk  Mon Nov 29 07:49:26 1999
From: S.I.Reynolds at cs.bham.ac.uk (Stuart Reynolds)
Date: Mon, 29 Nov 1999 12:49:26 +0000
Subject: discrete event simulation?
References: <xOe04.8702$cz3.315201@nnrp1.uunet.ca>
Message-ID: <38427656.5AEE@cs.bham.ac.uk>

Greg Wilson wrote:
> 
> Has anyone written a discrete event simulation framework in Python?  If so,
> I'd be grateful for pointers.

That depends upon what you mean. I have some classes for Markov Decision
Processes (of which Markov processes are a subset) and algorithms for
performing dynamic programming (value iteration) on them.

Stu



From mhammond at skippinet.com.au  Sat Nov 13 18:38:55 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sat, 13 Nov 1999 23:38:55 GMT
Subject: NT admin scripts with python
References: <000101bf2e22$7060a190$020a10ac@p350.gnaa.net>
Message-ID: <jEmX3.3257$Xo.17395@news-server.bigpond.net.au>

Yep - the win32net module is all you need!  Included in the win32all
extensions...

Mark.

geoff nordli wrote in message
<000101bf2e22$7060a190$020a10ac at p350.gnaa.net>...
>Does python have a module that would allow a person to write
>scripts on NT Server to lets say:  create users, map drives,
>modify registry, create shares, and so on.
>
>thanks,
>
>Geoff Nordli
>
>
>





From tbryan at python.net  Mon Nov 15 06:37:52 1999
From: tbryan at python.net (Thomas A. Bryan)
Date: Mon, 15 Nov 1999 06:37:52 -0500
Subject: GNU readline module
References: <80lcsk$99g$1@watserv3.uwaterloo.ca> <m3puxdpee2.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <382FF090.E7901DC2@python.net>

Michael Hudson wrote:

> > I have a nagging notion that maybe the readline
> > module is only for use from withing the
> > interactive interpreter.  IS this true? if so..
> > that's extremely dissapointing.>
 
> I've never seen it used for anything else. There's no particular
> reason for this. 

I've seen other people post about using it in their apps, and I 
don't think that there's any reason to use it only in the interpreter.
Of course, readline is GPL, right?  That might be unacceptable to 
those shipping closed, proprietary software.

---Tom



From gek8365 at garnet.acns.fsu.edu  Tue Nov 16 08:47:35 1999
From: gek8365 at garnet.acns.fsu.edu (Glenn Kidd)
Date: Tue, 16 Nov 1999 08:47:35 -0500
Subject: SMTP
Message-ID: <80rn4o$bnl$1@news.fsu.edu>

Does anyone know where some good info on Python's smtplib?  I have looked at
http://wwww.python.org but I was wondering if there was any more info out
there.  Any help would be appreciated.


Glenn





From guido at cnri.reston.va.us  Tue Nov  9 14:26:34 1999
From: guido at cnri.reston.va.us (Guido van Rossum)
Date: 09 Nov 1999 14:26:34 -0500
Subject: Fastest GUI toolkit?
References: <01bf29f1$cbfd5e60$5a41c3d1@Cyberus.cyberus.ca> <slrn82eapd.pct.nascheme@adler.dynodns.net>
Message-ID: <5lk8nr8o0l.fsf@eric.cnri.reston.va.us>

> Gordon Williams <g_will at cyberus.ca> wrote:
> >
> >I have an application where I need to draw many lines fast and repeatedly
> >(an oscilloscope).  The number of lines ranges from about 1000 to 16000. 
> >Currently I have the application using Tkinter but I am finding it too
> >slow.

nascheme at enme.ucalgary.ca (Neil Schemenauer) writes:

> I assume you are using the canvas to draw.  I think it is much
> faster to use Canvas.coords to move the lines then to redraw
> them.  If this approch works for your application it may be worth
> a try.

This may not save Gordon's app, but I once successfully saved a
canvas-using Tkinter app by bypassing most of the Tkinter Python code,
and calling Tk directly using <widget>.tk.call(...).  It takes a bit
of effort to generate the right Tcl syntax, but it can speed line
drawing up by an order of magnitude...

--Guido van Rossum (home page: http://www.python.org/~guido/)



From drek at MonsterByMistake.Com  Tue Nov  2 14:18:47 1999
From: drek at MonsterByMistake.Com (Agent Drek)
Date: Tue, 2 Nov 1999 14:18:47 -0500 (EST)
Subject: capturing stderr && stdout
Message-ID: <Pine.LNX.4.10.9911021350390.15744-100000@jazz.monsterbymistake.com>

Hullo,

I have searched the web and all the python documents/examples I could find
before posting here. My apologies if I have missed the obvious :) (I bought
two books as well that I am going through)

Alot of what I need to do with python depends on launching and controlling
a subprogram called hscript. It's a 'closed' application that is not a full
bodied language like python. I am hoping to launch one instance of hscript
write to it and keep its stdout && stderr.

Most of our work gets done with csh scripts like the following:
<csh_script>
#!/bin/csh -f

#csh stuff goes here

hscript -q $HIPDIR/$HIPFILE << ENDCAT >>&! $LOG
	# now I write hscript commands
	opcf /obj
	#etc
ENDCAT
exit
</csh_script>

here's a stripped down version of what I have thus far.

<python_script>
#!/usr/local/bin/python

import os, sys
HSCRIPT = "hscript"

p = os.popen("%s" % HSCRIPT, "w")
p.write("opcd /obj\n")
p.write("opadd geo\n")
p.write("opls\n")
p.write("quit\n")
</python_script>

This is great so far but I just need stderr from hscript (just in case it
crashes:) and stdout (to keep a neat log file)

This will need to work on *nix.

I'm trying to move to python as our main scripting language so that we can
reuse code and integrate a gui (tkinter) easily into our apps.

If you're interested in what hscript is you can visit http://www.sidefx.com/
and check out houdini.

thanks for any tips,

	=derek

Monster By Mistake Inc > 'digital plumber'
mailto:drek at monsterbymistake.com
http://www.interlog.com/~drek
"This mail was powered by Linux & Vim"




From fredrik at pythonware.com  Tue Nov 23 09:58:02 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 23 Nov 1999 15:58:02 +0100
Subject: Python books
References: <199911230613.HAA21179@dionysus.fw.cuci.nl> <383A9EF4.5D1E8986@callware.com>
Message-ID: <00a901bf35c3$2549ed80$f29b12c2@secret.pythonware.com>

Ivan Van Laningham <ivanlan at callware.com> wrote:
> If "Python Annotated Archives" is /F's book, it's already out as an
> eMatter book.  Visit www.python.org for details on how to get it. 

(or see below ;-)

well, the eff-bot guide is definitely an annotated script
collection, but I think Hans was referring to Martin Brown's
upcoming pMatter book; here's fatbrain's description of it:

http://www1.fatbrain.com/asp/BookInfo/BookInfo.asp?theisbn=0072121041&from=GTY867

in short: $50, 800 pages, available in december

(a wild guess is that this book will feature less scripts
and more annotations, but I haven't seen it.  but I've
heard rumours that it'll contain an eff-bot script or
two...)

</F>

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




From bwinton at tor.dhs.org  Thu Nov  4 20:22:41 1999
From: bwinton at tor.dhs.org (Blake Winton)
Date: Fri, 05 Nov 1999 01:22:41 GMT
Subject: Apply a function to each list member?
References: <816010E2456BD111A48700805FBBE2EEFDF6FC@ex-quebec-u1.baan.com>
Message-ID: <slrn824gee.1f7.bwinton@tor.dhs.org>

On Fri, 22 Oct 1999 07:52:15 -0400, Gaetan Corneau wrote:
>I knew I had seen a function to do this somewhere, but I could not find it.
>"map" is not an intuitive name for me (maybe it would if english was my main
>language).

Speaking as someone who's one language is English, "Probably not".
"map" is really a Lisp/Scheme concept, and bears only a passing
relationship to English.  (Huh, what do ya know, looking up "map"
on http://www.m-w.com/ reveals (in meaning 4) that "map" actually
means "FUNCTION".  I have no idea why they capitalized it.)

>"Profanity is the one language all programmers know best"

Well, okay, maybe two languages.  ;)

Later,
Blake.

-- 
One Will.  One Dream.  One Truth.  One Destiny.  One Love.



From nemeth at iqsoft.hu  Thu Nov  4 16:05:57 1999
From: nemeth at iqsoft.hu (Németh Miklós)
Date: Thu, 04 Nov 1999 22:05:57 +0100
Subject: Extending Python dynamically
References: <012801bf26fb$b8e0a4c0$1a25d2d1@jenkondev.com>
Message-ID: <3821F535.D528D1BF@iqsoft.hu>

Robin Dunn wrote:

> > Thank you for your help. The method you suggested works perfectly!
> > However, I do not understand completely how the Python interpreter finds
> out
> > that it has to load an so file. Normally -- I believed -- when Python sees
> an
> > "import xxx" statement it looks for a xxx.py, xxx.pyc or xxx.pyo file in
> the
> > $PYTHONPATH directories or an xxx subdirectory having an __init__.py file,
> but
> > not an xxx.so file. This means that if I add the current directory to the
> > $PYTHONPATH (such as export PYTHONPATH='.' Python will find any arbitrary
> so
> > file in the $PWD?
>
> It will also load xxxmodule.so as well, but .so files only get imported if
> they have an initxxx() function.
>
> --
> Robin Dunn
> Software Craftsman
> robin at AllDunn.com
> http://AllDunn.com/robin/
> http://AllDunn.com/wxPython/  Check it out!

Robin, you are a Genius!  Neither the "Python Reference Manual Release
1.5.2's
section 6.11 "The import statement" nor the Extending and Embedding the
Python
Interpreter documents explain this behaviour of the Python interpreter.
That is
why I felt myself a bit lost.
But now with Your help, I feel I know "everithing" ;-) about Python
extensibility.
I am also very happy with this mailing list (another very good ones
where I am a member are Qt's, LyX's, Apache-SSL's, etc lists). Within
some hours I got all the
support I needed -- no commercial mega-corporations are able to provide
such a quality
support even for tons of money.  I remember when I spent three weeks
with MS
Visual J++ 6.0 and no one could answer my simplest questions: that was
my last
attemp to fight with a proprietary (especially MS) development tool.

Thanks again!

NM




From wayne_strauss at mail.medscape.com  Tue Nov  9 09:55:18 1999
From: wayne_strauss at mail.medscape.com (wayne_strauss at mail.medscape.com)
Date: Tue, 09 Nov 1999 14:55:18 GMT
Subject: Newbee WinNT ODBC Question
Message-ID: <809ckk$vbf$1@nnrp1.deja.com>

I've installed Python and Win32all.exe. When I run the following (only
the first three lines shown):

import dbi, odbc     # ODBC modules
import time          # standard time module

dbc = odbc.odbc('db/id/pw')

I get:

>>> Traceback (innermost last):
  File "d:\Program
Files\Python\Pythonwin\pywin\framework\scriptutils.py", line 237, in
RunScript
    exec codeObject in __main__.__dict__
  File "D:\dev\Python\ODBC\odbc.py", line 1, in ?
    import dbi, odbc     # ODBC modules
  File "odbc.py", line 4, in ?
    dbc = odbc.odbc('db/id/pw')
TypeError: call of non-function (type module)

If the import worked then why does the actual function call fail?


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



From justus at my-deja.com  Mon Nov 29 15:34:37 1999
From: justus at my-deja.com (Justus Pendleton)
Date: Mon, 29 Nov 1999 20:34:37 GMT
Subject: predicate classes
Message-ID: <81uo0t$2rr$1@nnrp1.deja.com>

I decided to broaden my horizons so I was reading about/playing around with
Cecil.	One of the things Cecil has that I hadn't heard of before are
predicate classes
(http://www.cs.washington.edu/research/projects/cecil/cecil/www/www/Papers/pr
edicate-classes.html), which I thought were nifty beyond all belief.

I also thought it would be nifty if Python had them so I got the source
from the CVS respository and started hacking the grammar and realized I
had no idea what I was doing :-).

Before getting in any deeper in something I only vaguely understand I
thought I would ask the opinions of people who know more about Python
than I whether they think it would be
easy/hard/well-damn-nigh-impossible to add this kind of dynamic retyping
of objects based on predicates.

(I did a quick search on deja and didn't find any talk about this, so
apologies in advance if this is something that has been as talked to
death as the indentation thing.)


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



From jblazi at netsurf.de  Tue Nov 16 09:10:37 1999
From: jblazi at netsurf.de (Janos Blazi)
Date: Tue, 16 Nov 1999 15:10:37 +0100
Subject: Beginning programming with Python
References: <sng0ychgwt.fsf@motorola.com> <382B3A53.5816B2@callware.com> <382B3D79.74CD3E0F@email.sps.mot.com> <80p9ds$ppo$1@newshost.accu.uu.nl> <asteroidal-80psfo/INN-2.2.1/bat@broadway.news.is-europe.net> <38317787.19AFDAF5@compaq.com>
Message-ID: <becket-80rol2/INN-2.2.1/bluejacket@broadway.news.is-europe.net>

I agree with that.
Janos B.

Greg Ewing <greg.ewing at compaq.com> schrieb in im Newsbeitrag:
38317787.19AFDAF5 at compaq.com...
> Janos Blazi wrote:
> >
> > the indenting this time? There is actually nothing else you can forget
in
> > Python.
>
> Well, not quite - you can forget the "self" in method
> declarations fairly easily...
>
> Greg





From emile at fenx.com  Thu Nov  4 09:52:11 1999
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 4 Nov 1999 06:52:11 -0800
Subject: __str__ vs. __repr__
References: <000401bf25b9$88ea57e0$2e2d153f@tim>  <199911040157.UAA11908@eric.cnri.reston.va.us>
Message-ID: <046101bf26d4$2dd1a720$01ffffc0@worldnet.att.net>

----- Original Message -----
From: Guido van Rossum <guido at cnri.reston.va.us>
To: Tim Peters <tim_one at email.msn.com>
Cc: <python-list at python.org>
Sent: Wednesday, November 03, 1999 5:57 PM
Subject: Re: __str__ vs. __repr__
<snip>
> Any suggestions?
>
> --Guido van Rossum (home page: http://www.python.org/~guido/)


It seems to me that issues arise when using the interactive interpreter.
I suspect that little useful work is done in this mode, and is used
predominately for quick testing while creating python programs.

Also, as entering a value without the print is a shortcut/convenience
used during testing, why not supply even more information?  Supply the
class, type, lengths, indexes, nested, pretty-printed, etc.  Use the
repr and str if it helps.  Interactive mode is for learning, testing,
and experimenting.  Let it be of as much help as possible.

I-type-print-so-fast-sometimes-it-comes-out-printprint-ly yr's


Emile van Sebille
emile at fenx.com
-------------------







From gmcm at hypernet.com  Mon Nov 29 12:55:35 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Mon, 29 Nov 1999 12:55:35 -0500
Subject: Access violation while trying to replace stdout with a local disk file
In-Reply-To: <81udcl$nm3$1@news1.sinica.edu.tw>
Message-ID: <1268224356-30441862@hypernet.com>

SINICA writes:
> 
> I'm trying to redirect the output of Python into a external
> diskfile since my program and python interpretor are two
> different processes (python15.dll and myprog.exe)
> PyObject_Print() will cause a "Access Violation" while trying to
> print on the file handle which is created by myprog.exe
> 
> any suggestion will be greatly appreciated.

They both need to be linked to the same c-runtime library, (ie, 
"Multithreaded DLL" in Settings | Compile | Code Generation).

- Gordon



From herzog at online.de  Fri Nov  5 06:53:51 1999
From: herzog at online.de (Bernhard Herzog)
Date: 05 Nov 1999 12:53:51 +0100
Subject: __str__ vs. __repr__
References: <000401bf25b9$88ea57e0$2e2d153f@tim> <199911040157.UAA11908@eric.cnri.reston.va.us> <m31za6339e.fsf@greebo.nodomain.de> <5logd98v8z.fsf@eric.cnri.reston.va.us>
Message-ID: <m3ln8di27k.fsf@greebo.nodomain.de>

Guido van Rossum <guido at cnri.reston.va.us> writes:

> Bernhard Herzog <herzog at online.de> writes:
> 
> > How about a hook function a la __import__? I.e. have a function
> > __format__ [1] in __builtin__ that is bound to repr by default. Whenever
> > Python prints the result of an expression in interactive mode, it calls
> > __format__ with the result as parameter and expects it to return a
> > string ready for printing.
> 
> Brilliant!

Glad you like it!

>  The default function could do all the current built-in
> magic -- print nothing if it's None, assign it to builtin _, and so
> on.

It seems to me that assigning to _ and printing the result are two
separate things. IMO, the hook function should only do the formatting,
the assignment could still be hardwired in the interpreter loop. Pseudo
code:

while 1:
	statement = read_statement()
	if not statement:
		break
	result = run_statement(statement)
	__builtin__._ = result
	if result is not None:
		print __format__(result)


> I wonder if it would have to have a __magic__ name?  It could be
> called display and it could be documented and usable in
> non-interactive programs as well.  Or am I getting carried away?
> (Possibly the habit if assigning to _ and suppressing None would make
> the default display() function a bit clumsy to use.)

Well, I chose a magic name because it's called in special circumstances.
Calling it display is fine with me, especially if it really does only
formatting and no assignment to _ and no special-casing of None, as is
assumed in my pseudo code above. Such a version of display would also be
usable in non-interactive programs.

One could take this even further by moving the entire interactive mode
to a python module. Most of the python code for that is already
available in the code.py module, IIRC.

-- 
Bernhard Herzog	  | Sketch, a drawing program for Unix
herzog at online.de  | http://www.online.de/home/sketch/



From kelley at bioreason.com  Tue Nov  2 13:55:11 1999
From: kelley at bioreason.com (Brian Kelley)
Date: Tue, 02 Nov 1999 11:55:11 -0700
Subject: cStringIO versus StringIO: "readlines"
Message-ID: <381F338E.D3D2E748@bioreason.com>

I noticed that StringIO has a "readlines" member function and cStringIO
doesn't.  Is anyone planning on adding this in the near future?  I was
wondering if I should go ahead and do it or if there is against.

The cStringIO documentation just says "provides an interface similar to
StringIO" and StringIO says "see the description on file objects."
Needless to say, I use readlines() quite a bit.

--
Brian Kelley          w 505 995-8188
Bioreason, Inc        f 505 995-8186
309 Johnson Av
Santa Fe, NM 87501    kelley at bioreason.com






From junkster at nospam.rochester.rr.com  Thu Nov 25 20:53:34 1999
From: junkster at nospam.rochester.rr.com (Benjamin Schollnick)
Date: Fri, 26 Nov 1999 01:53:34 GMT
Subject: Journaling Code
References: <81k6m9$gkm$1@newsie2.cent.net> <19991125174544.A1164@better.net>
Message-ID: <rNHvjEdhm5Pp-pn2-WvJC9LKJudGl@d185d1884.rochester.rr.com>

On Thu, 25 Nov 1999 22:45:44, William Park <parkw at better.net> wrote:


> Not coming from 3D rendering background... but, wouldn't it be easier
> to have your modeller produce "log" file of anything coming in?

Honestly, IMHO, a log file (for that use) is basically a "poorman's"
journal.

		- Benjamin

====================================================
          (Remove "NoSpam" to Email me)
====================================================
Please feel free to copy any and or all of this sig.
A little something for spam bots:

root at localhost postmaster at localhost admin at localhost
abuse at localhost postmaster at 127.0.0.1

Chairman William Kennard: bkennard at fcc.gov 
Commissioner Susan Ness: sness at fcc.gov
Commissioner Harold Furchtgott-Roth: hfurchtg at fcc.gov
Commissioner Michael Powell: mpowell at fcc.gov
Commissioner Gloria Tristani: gtristan at fcc.gov
consumerline at ftc.gov
fccinfo at fcc.gov
ssegal at fcc.gov




From volumeregeling at hotmail.com  Mon Nov 29 12:53:45 1999
From: volumeregeling at hotmail.com (volumeregeling)
Date: Mon, 29 Nov 1999 17:53:45 GMT
Subject: Need help with html files
Message-ID: <3842bd0e.23017938@122.0.0.250>

hi,

i would like to open a local html file from within python,
When the html is opened, it should be opened like if you double click
on the filename, so it uses the know filetype from windows.

in short , when someone has netscape as the default browser.
I open the html file python should call the html file , and windows
opens the default browser with the file as argument.

I hope this is clear enough, at the moment i can?t think of a way to
explain it differently.

if you know what i mean and if you have a solution for my problem
please post or mail.

thanks





From tim_one at email.msn.com  Fri Nov 26 21:32:46 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Fri, 26 Nov 1999 21:32:46 -0500
Subject: python test frameworks
In-Reply-To: <slrn83u3um.4ge.quinn@groat.ugcs.caltech.edu>
Message-ID: <000401bf387f$b12d47c0$742d153f@tim>

[Quinn Dunkan]
> ...
> I'm looking for some information on techniques for self-testing
> pyhon code (something I should have been doing before, but only
> now am I writing a system complex enough to warrant it).

I would have thought you had written at least one function before now <0.9
wink>.

> I remember Tim's thread about his doc-string embedded testing, and
> that looked interesting, but I can't find it now.

http://www.python.org/ftp/python/contrib/System/doctest.README

if-it's-not-tested-it's-broken-ly y'rs  - tim





From af137 at torfree.net  Sun Nov 28 05:16:02 1999
From: af137 at torfree.net (Al Aab)
Date: Sun, 28 Nov 1999 10:16:02 GMT
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <38397696@news.xtml.co.uk> <221119992150405215%zippy@cs.brandeis.edu> <383a8909@news.xtml.co.uk> <slrn840dcf.bt2.wtanksle@hawking.armored.net>
Message-ID: <FLwKIr.6wn.0.queen@torfree.net>

some text editors take care/ease LISP perens pain.
emacs is one.

over a decade ago
there was a lisp written for TOPS10  os;

its editor used to close the remaining parens for u 
(just hit ESC ESC)
was really neat.

even emacs cannot do that,
unless u program it in emacs lisp.
-- 
=-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-
al aab, ex seders moderator                                   sed u soon 
               it is not zat we do not see the  s o l u t i o n          
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-+



From tim_one at email.msn.com  Wed Nov 24 16:00:31 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 24 Nov 1999 16:00:31 -0500
Subject: Python complaints
In-Reply-To: <005201bf367e$4901fef0$f29b12c2@secret.pythonware.com>
Message-ID: <000601bf36be$f1bf5e80$292d153f@tim>

[Thomas Hamelryck]
> I don't want to start the discussion about indentation again;

Heh heh -- good one.

> it's no use since indenting blocks is here to stay. The point is that
> if you want to write code that looks ugly or unreadable you will
> succeed in doing that, indentation or not.  In many cases, using a
> variable indentation (depending on what construct you are using or
> the length of a line) leads to code that is visuably much more appealing.

[/F]
> you make that sound like a scientific fact.  can
> you point us to some research, or is it just yet
> another random opinion?

I suspect Thomas has the same problem with citing research as I have:  my
software imposes a 2Mb limit on msgs I can post <wink>.

There is one (& only one) area where Python's indentation gets in my way a
bit:  conceptual blocks that can't be written as blocks.  These typically
occur when toggling a mode or flag of some sort, like

    mutex.acquire()
    do stuff
    mutex.release()

or

    myIEEEstuff.set_rounding_mode(CHOP)
    do stuff
    myIEEEstuff.restore_rounding_mode()

In C++ it's idiomatic to write a class that sets the mode in the constructor
and restores it in the destructor, then just declare a temp object of that
class in a new block; e.g.,

    {
        CriticalSection guardme;  // ctor acquires, dtor releases
        do stuff
    }
    // the language guarantees the dtor has been called at this point

This is very powerful, and the "extra" indentation is helpful in pointing
out that something odd is going on.  Alas,

    mutex.acquire()
        do stuff
    mutex.release()

is a syntax error in Python.  The closest I can get is

    mutex.acquire()
    try:
        do stuff
    finally:
        mutex.release()

which is exactly the semantics I want, but if "do stuff" is more than a
couple of lines strains my pattern-recognition wetware.

Fortunately, this is so infrequent an irritation that I won't even mention
it.

not-responding-to-a-discussion-that-never-started-ly y'rs  - tim





From akuchlin at mems-exchange.org  Tue Nov 30 14:36:34 1999
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: 30 Nov 1999 14:36:34 -0500
Subject: XML tool status?
References: <DIv04.8846$cz3.320849@nnrp1.uunet.ca>
Message-ID: <3dhfi3hiv1.fsf@amarok.cnri.reston.va.us>

"Greg Wilson" <gvwilson at nevex.com> writes:
> 2) the Python XML SIG status page has a "most recent update" date from
> December 1998 --- is that fossilware, or has there really not been anything
> new posted since then?

The code's been developed subsequently, and the CVS tree is much more
recent.  I've simply been remiss in updating the Web pages (and in
getting a 0.5.2 release out).

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
Where is human nature so weak as in the bookstore?
    -- Henry Ward Beecher




From mwh21 at cam.ac.uk  Mon Nov  8 14:39:38 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 08 Nov 1999 19:39:38 +0000
Subject: pythonistas or pythoneers?
References: <Pine.LNX.4.10.9911081026470.4450-100000@robert.polytopic.com> <slrn82e4mp.43f.thor@localhost.localdomain> <8076ed$d2n$1@nnrp1.deja.com>
Message-ID: <m39048ojr9.fsf@atrus.jesus.cam.ac.uk>

init__self at my-deja.com writes:

> > So pythonistas sounds extremely powerfull, because python can
> > be spelled in Spanish as "piton", that means "horn" ( like the bull's
> > horns). So pythonistas sounds like the "damned crazy followers
> > of the horns" :DD
> 
> So instead of those masturbatory "Just another you-know-what hacker"
> .sig's, one could have
> 
>    #!/usr/local/bin/python
>    print "Just another damn crazy follower of the horns."
> 
> as a fine .sig :-) It conveys succinctly the readability and simplicity
> of the language!

I *like* that! Or maybe:

#!/usr/local/bin/python
print "I can read what I write"
print "I can even read what you write"

though that might be considered offensive in some circles...

Regards,
Michael



From sai at sw.ods.com  Wed Nov 10 10:54:10 1999
From: sai at sw.ods.com (Sai Ramanath)
Date: Wed, 10 Nov 1999 09:54:10 -0600
Subject: Exceptions: ^z
Message-ID: <38299522.5D06E8D5@sw.ods.com>

Is there a way to catch Ctrl-Z which in UNIX lets the user suspend the Python
shell into background?

I know we can catch Ctrl-C and Ctrl-D using KeyboardInterrupt and EOFError
respectively.

Thanks,
Sai
-------------- next part --------------
A non-text attachment was scrubbed...
Name: sai.vcf
Type: text/x-vcard
Size: 362 bytes
Desc: Card for Sai Ramanath
URL: <http://mail.python.org/pipermail/python-list/attachments/19991110/4984e920/attachment.vcf>

From emil at imf.au.dk  Mon Nov  8 10:16:25 1999
From: emil at imf.au.dk (Emil Hedevang Lohse)
Date: Mon, 08 Nov 1999 16:16:25 +0100
Subject: Graphics.
Message-ID: <3826E949.BBA1BAFD@imf.au.dk>

How do I draw some graphics on a screen using Python running on
HP-UX
MacOS
Windows

???

>From Emil
emil at imf.au.dk



From g_will at cyberus.ca  Mon Nov 15 15:22:16 1999
From: g_will at cyberus.ca (Gordon Williams)
Date: Mon, 15 Nov 1999 20:22:16 GMT
Subject: Line Benchmarks Tkinter vs. wxPython
Message-ID: <01bf2fa7$02992ca0$8a42c3d1@Cyberus.cyberus.ca>

Hello All,

I have done some benchmark testing to see the difference in speed between
plotting multi point lines using Tkinter and wxPython.  I have looked at 3
different methods for doing it in Tk and 2 different methods in wx.  The
differences are astounding for the 1000 to 10000 point lines that I
examined.  Generally, wx was found to be in the order of 200 to 1800 times
faster than the standard Tkinter.  Tk can be sped up by ten to 50 times by
changing Tkinter so it bypasses the _flatten method or by making the Tk
calls directly.

A repeating sine wave was created and put into a horizontal scrollable
window of about 500 points wide and 300 points high.  These are the results
for the line plotting time only.

Time in Seconds on a P166 with 96MB RAM.

Points		Tkinter		Tkinter		Tkinter		wxPython	wxPython
(standard)	(mod 1)	(mod 2)	(direct draw)	(bitmap)
1000		1.190		0.192		.202		0.007		0.026
2000		4.406		0.385		--		0.014		0.042
5000		25.170		0.967		--		0.027		0.080
10000		100.260	1.941		2.064		0.054		0.148

Notes:
1)	Tk standard is the standard usage with no modifications.

2)	Tk mod 1 has Tkinter modified so a flat tuple of points is passed to
create_line and _flatten is by-passed.

3)	Tk mod 2 used the apply method  e.g. 
apply(self.can.tk.call,(self.can._w,'create','line') + points + 
('-width',1,'-fill',colour,'-tags',"tag"))

4)	wx direct draw draws to the window directly.

5)	wx bitmap draws to a bitmap and then blit



Conclusions:

1)	The time to plot lines using the standard Tkinter is quadratic with the
number of points (ugh!).  This is due to the time spent in _flatten. 
_flatten is very versatile but agonizingly slow for large numbers of lines.

2)	Tkinter mod 1 and 2 give the same results.  Times are linear with the
number of points plotted as both cases dont use the _flatten method.  This
means that a flat tuple of points has to be used (a small constraint).  The
speed increase in doing this is from 10 to 50 times over the standard.

3)	wxPython beats Tk by a very! large margin even when Tk is called
directly.  Direct drawing onto to screen is roughly 3 times faster than
using a bitmap.



Thanks to:
Greg Landrum for providing some information on direct calls to tk.
Ionel Simionescu for providing the wxPython code which I was able to
modify.




From phd at phd.russ.ru  Fri Nov 19 08:28:57 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Fri, 19 Nov 1999 13:28:57 +0000 (GMT)
Subject: urllib.getsize()
In-Reply-To: <19991119133404.A3308@optiplex.palga.uucp>
Message-ID: <Pine.LNX.4.21.9911191327560.13314-100000@fep132.fep.ru>

On Fri, 19 Nov 1999, Gerrit Holl wrote:
> Another problem: does every HTTP server give the Content-Length header?

   Certainly NO. Most CGIs do not set the header. Often server-side
includes loose the header.

Oleg.
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From greg.ewing at compaq.com  Fri Nov 19 10:27:48 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Sat, 20 Nov 1999 04:27:48 +1300
Subject: Type/Class Unification in Ruby
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp> <7vjtrc$hmm$1@newshost.accu.uu.nl> <38204614.2EAC6051@compaq.com> <5ld7trtak5.fsf@eric.cnri.reston.va.us> <3826D63C.2731BB1F@compaq.com> <38328948.9E35556F@compaq.com> <m3d7t9v8lr.fsf@qiao.localnet>
Message-ID: <38356C74.F9E4248D@compaq.com>

Clemens Hintze wrote:
> 
> But it is worth mentioning, that Ruby features
> single inheritance *by*purpose*. Not because of the implementation
> details!

I wasn't making a value judgement, only pointing out that
Ruby's solution is as simple as it is because of this
restriction. 

> If it would implement multi-inheritance, the class structure could
> e.g. use an array of pointer to all such C structs, not only one
> pointer, as implemented now!

Sure, Python could do something like that too. But it would
get very hairy and complicated and probably slow everything
down a lot, casting doubt on whether it's worth the effort.

A middle ground might be to allow inheriting from any number
of user-defined classes, but at most one built-in type.

> The second (preferred) way is, to build instances of class DATA
> wrapping the C struct you have allocated for your extension.

Yes, I know, but what I was saying still applies in that case.
All the methods of the extension class expect the DATA object
to be pointing to their own particular kind of struct, and
because of the inheritance structure, this is always true --
so they don't need to bother checking.

> Deriving a Ruby class coded in C from another one also coded in C,
> seems to be the only right tricky part of that language, IMHO. :-)

I think it will work as long as the derived class makes sure
that its C structure is an exension of the superclass's C
structure.

>     Greg> [global instance var dict] might cause ref counting
>     Greg> problems.
> 
> Once again, thanks for Ruby's true garbage collection! :-)

Garbage collection alone doesn't actually solve the problem;
the references from the global dict need to be treated as
weak references somehow under either scheme. I haven't
looked, but presumably Ruby must be ignoring the global dict
refs when marking, and cleaning them up whenever a builtin
or extension object is deallocated. Python could be made to
do the equivalent.

> The big exception,
> AFAIK, was class Fixnum. It never had any instanciation method, as it
> doesn't wrap any struct.

Yes, that's the one class you really can't instantiate a
subclass of, since its class is totally implicit.

But it's not clear whether it's any use having something
like a UserFlonum, either, even though technically it could
be done. Unless you override every method, any arithmetic
on it would return an ordinary Flonum. And if you are going
to override every method, it might as well be a totally
new class!

> Often it is better to use the module
> mechanism to add methods to class Array or some of its instances
> during runtime.

You mean you can add methods to an *existing* class that
way? Not sure if I like that idea...

Greg



From alan.gauld at gssec.bt.co.uk  Mon Nov 29 12:49:24 1999
From: alan.gauld at gssec.bt.co.uk (Alan Gauld)
Date: Mon, 29 Nov 1999 17:49:24 +0000
Subject: yet 'nother newbie question
References: <3840300F.66DC46DD@wjk.mv.com>
Message-ID: <3842BCA4.51B15166@gssec.bt.co.uk>

William J. King wrote:
> When passing an argument to a function etc. -- how does one determine
> that it is
> an integer that is being passed and not a string for example... 

Another poster answered this.

However be careful that your not trying to use Python like C++
Static typing is useful in C++ but dynamic typing is much 
better for Python's typical applications. More flexible and 
faster to develop. Rather than using try/except to test types 
use it to catch the few occasions when a bad typ[e is passed.

Your fellow programmers are not totally stupid and will by 
and large try to pass the right thing. If you catch the 
errors it usually works...

If you do need to check the type then type() is your friend.

Alan G.

-- 
=================================================
This post represents the views of the author 
and does not necessarily accurately represent 
the views of BT.



From a.eyre at optichrome.com  Wed Nov 24 04:33:38 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Wed, 24 Nov 1999 09:33:38 -0000
Subject: Embedding: Getting the module from a PyObject *
In-Reply-To: <E11qKPm-0004bF-00@rzstud1.rz.uni-karlsruhe.de>
Message-ID: <002a01bf365e$fcc3cf60$3acbd9c2@peridot.optichrome.com>

> can anyone tell me how to get the module object that a certain instance's
> class is defined in, from that instance?

Assuming:

class Spam:
    pass

s = Spam()

Then:

s.__class__ returns class Spam

Spam.__module__ or s.__class__.__module__ returns the module it was defined
in

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com
--------------------------------------------




From stuart.hungerford at webone.com.au  Wed Nov  3 03:15:12 1999
From: stuart.hungerford at webone.com.au (Stuart Hungerford)
Date: Wed, 03 Nov 1999 19:15:12 +1100
Subject: How can I make my assertions smarter?
References: <000601bf25be$0cdeb880$2e2d153f@tim>
Message-ID: <381FEF10.3B27C05B@webone.com.au>

Tim Peters wrote:

> [Stuart Hungerford, desperately hoping that design by contract will
>  be easier in Python if he calls it by some other name ...]

   I'd kind of given up on true DBC, and just went with loads of asserts
   as ersatz pre and postconditions. When I find myself calling the same
   construct hundreds of times, I can't help but feeling I may be able to
   do things either more conveniently, more flexibly or with more value
   added.  Hence my question.

> > I'm using some simple functions to implement crude precondition,
> > postcondition and invariant checking in Python classes.
> >
> > (Tim: please be gentle ;-)
>
> I am always gentle with you, Stuart!  Your heart is in the right place --
> although it may as well be firmly stuck up your ass for all the good it's
> doing you <wink>.

   This must be the Python equivalent of "tough love" ;-)   Top of my
   new years resolutions list is not to post any more DBC in Python
   questions.

   Thanks to everyone who responded to this question -- I've learnt some
   interesting and useful techniques in the process.

Ciao

Stu






From ajung at sz-sb.de  Tue Nov 16 01:19:38 1999
From: ajung at sz-sb.de (Andreas Jung)
Date: Tue, 16 Nov 1999 07:19:38 +0100
Subject: list.without()?
In-Reply-To: <y0jln7zi9pb.fsf@vier.idi.ntnu.no>; from mlh@vier.idi.ntnu.no on Tue, Nov 16, 1999 at 01:03:44AM +0100
References: <y0jln7zi9pb.fsf@vier.idi.ntnu.no>
Message-ID: <19991116071937.A23463@sz-sb.de>

On Tue, Nov 16, 1999 at 01:03:44AM +0100, Magnus L. Hetland wrote:
> 
> Hi!
> 
> Is there a function or method in the standard language that can
> non-destructively return a list without a specified element? Would it
> be possible to add it as a method to lists?

mylist = filter(lambda x: x != ItemToRemove , mylist)

or something similar .... 



From mwh21 at cam.ac.uk  Mon Nov 15 14:35:29 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 15 Nov 1999 19:35:29 +0000
Subject: exec/eval changed?
References: <942690155.686795566@news.glph1.on.wave.home.com>
Message-ID: <m34sen35vi.fsf@atrus.jesus.cam.ac.uk>

JB <jimbag at kw.igs.net> writes:

> Did the exec/eval statements change somehow in the 1.5 versions? I
> have a script that did an eval( "some code", dict1, locals()) but
> now it doesn't work. I've looked through the doc's but they are
> little thin as to these statements. Any help is appreciated.

Well, eval("some code") is a syntax error. What are you trying to to
do? We can't read minds here.

Regards,
Michael



From fredrik at pythonware.com  Wed Nov 17 08:00:03 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 17 Nov 1999 14:00:03 +0100
Subject: Need help dealing with spaces in NT paths.
References: <80pr76$jq5$1@nnrp1.deja.com> <11f733ec.96958819@usw-ex0101-007.remarq.com> <uogcu93dt.fsf@cs.uu.nl> <80sj4q$l7t$1@nnrp1.deja.com> <80tgi8$hbn$1@ssauraaa-i-1.production.compuserve.com> <002501bf30e8$a41a5bc0$f29b12c2@secret.pythonware.com> <001301bf30f8$fa61cf80$224dc0d8@josefbet>
Message-ID: <007301bf30fb$aa921c50$f29b12c2@secret.pythonware.com>

> >>> os.system('dir "program files"')
> 
>  Volume in drive C is CDRIVE
>  Volume Serial Number is 2E50-0FE9
>  Directory of C:\WINDOWS

are you sure you have a "program files" directory in
your "windows" directory?  it's usually placed under
the root directory...

</F>




From kyriacou at umbc.edu  Wed Nov 24 14:33:57 1999
From: kyriacou at umbc.edu (Stelios Kyriacou)
Date: 24 Nov 1999 14:33:57 -0500
Subject: More problems with random seed (was Re: whrandom seed question)
References: <81eu22$411al$1@umbc7.umbc.edu> <m366ysu7t4.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <81hej5$3tmv0$1@umbc7.umbc.edu>

Thanks both for your correct answer. Unfortunately I hate to say but my
ultimate question (which I never mentioned :) ) is unanswered. Why
when I call the gauss distribution I get 2 alternating values:

>>> random.seed(0); print random.gauss(0,1); 
2.11224018067
>>> random.seed(0); print random.gauss(0,1); 
0.225551631579
>>> random.seed(0); print random.gauss(0,1); 
2.11224018067
>>> random.seed(0); print random.gauss(0,1); 
0.225551631579
>>> random.seed(0); print random.gauss(0,1); 
2.11224018067
>>> 


Any more ideas? 
Stelios




From bmcd at es.co.nz  Tue Nov 23 15:49:23 1999
From: bmcd at es.co.nz (Ben Caradoc-Davies)
Date: 23 Nov 1999 20:49:23 GMT
Subject: Method objects question
References: <slrn83ku3j.2je.bmcd@ripley.sulaco.net> <slrn83krsu.2i0.bmcd@ripley.sulaco.net> <383a7309@news.uni-bielefeld.de> <81eha1$rf6$1@nntp6.u.washington.edu>
Message-ID: <slrn83lvr7.e7.bmcd@ripley.sulaco.net>

On 23 Nov 1999 17:01:53 GMT, Donn Cave <donn at u.washington.edu> wrote:
>It doesn't seem to matter where a function is defined, only the scope
>where it's found.  The normal and most convenient way to get a function
>into the class name space is certainly to define it there, but any other
>way will serve as well - it doesn't matter if you define it outside the
>class, or inside.
>How could we justify anything different, in a dynamic "run time" system
>like Python?

Sure, the class behavious is excellent. But the point is, why should functions
found in a class namespace be converted into methods, but functions found in an
instance namespace not converted?

For example, suppose you have a running application containing two instances of
the same class. You need to patch your application on the fly, performing
*different* modifications on each object. You send a signal to your
application, and when it is next idle it execfiles your patch script.

This script can modify each instance's variables, rebinding them to entirely
new objects. In particular, it is possible to hide class variables by adding
instance variables of the same name. However, with the current behaviour, it is
not possible to bind instance variables to function objects and override class
variables with the same name (bound to function objects) without changing any
other code, because the rest of the program will expect these instance
atributes to be *method objects*, not ordinary function objects. It is only
possible to change the class variable, and thus change the behaviour of *all*
instances of that class.

In a nutshell, I don't see any fundamental reason why the
search-instance-namespace then search-class-namespace attribute resolution
order couldn't be used to allow method overriding on an instance-by-instance
basis. With the current, function-to-method conversion rules, this is
impossible.

Actually, I'm more curious about why this design decision was made. (Or did it
just happen that way!)

-- 
Ben Caradoc-Davies <bmcd at es.co.nz>



From Vladimir.Marangozov at inrialpes.fr  Mon Nov 29 01:17:08 1999
From: Vladimir.Marangozov at inrialpes.fr (Vladimir Marangozov)
Date: Mon, 29 Nov 1999 07:17:08 +0100
Subject: AIX Py compiling errors?
References: <81r96t$meq$1@nnrp1.deja.com>
Message-ID: <38421A64.6A7001E2@inrialpes.fr>

moonseeker at my-deja.com wrote:
> 
> Hi!
> 
> Anyone using Python on AIX here? I tried to compile Python on my AIX 4.1.5
> box using the IBM compiler but I get an error (or warning):
> 
>   cc_r  -O -qmaxmem=4000 -I./../Include -I.. -DHAVE_CONFIG_H -c
> ./socketmodule.c "./socketmodule.c", line 1127.22: 1506-280 (W) Function
> argument assignment between types "struct sockaddr*" and "unsigned char*" is
> not allowed.
> 
> Anyone a suggestion?

A warning. It doesn't hurt.

-- 
       Vladimir MARANGOZOV          | Vladimir.Marangozov at inrialpes.fr
http://sirac.inrialpes.fr/~marangoz | tel:(+33-4)76615277 fax:76615252



From andres at corrada.com  Sun Nov 21 10:28:18 1999
From: andres at corrada.com (Andres Corrada)
Date: Sun, 21 Nov 1999 10:28:18 -0500
Subject: Order of keyword arguments
References: <38379233.ADE65CDB@corrada.com> <19991121034245.A21255@quark.emich.edu>
Message-ID: <38380F92.D9817451@corrada.com>

Jeff wrote:

> 
> keyword arguments are passed as a dictionary, and thus the 'order' of the
> arguments will not be preserved. what is the advantage you see to being able
> to obtain that information? wouldn't a redesign of the interface keep things
> simpler and increase the likelihood of success?
> 

I want to have my cake and eat it too! I thought this would be a way of
keeping the flexibity of keyword arguments but also pass information
along on how the calling function wants the items returned. For example,
an HTML form could be called with a variable number of input tags and
you want to render the HTML in the same order that they were used to
call the function.

------------------------------------------------------
Andres Corrada-Emmanuel   Email: andres at corrada.com
Owner                     http://www.corrada.com/mamey
Mamey                     Phone: (413) 587-9595
------------------------------------------------------



From mwh21 at cam.ac.uk  Fri Nov 26 08:49:11 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 26 Nov 1999 13:49:11 +0000
Subject: dynamic modules?
References: <000101bf37fe$94dac790$3acbd9c2@peridot.optichrome.com>
Message-ID: <m3iu2ps6qw.fsf@atrus.jesus.cam.ac.uk>

"Adrian Eyre" <a.eyre at optichrome.com> writes:

> > Is there a chance to write modules I can dynamically change? I tried
> > reload(), but thats not working because I can't reload modules imported
> > with 'from XXX import *'.
> 
> # Reload any/all modules (N.B. untested)
> import sys
> 
> def reloadAll():
>     for mod in sys.modules.keys():
>         reloadModule(mod)
> 
> def reloadModule(name):
>     reload(sys.modules[name])

equivalently:

def reloadAll():
    for mod in sys.modules.values():
        reload(mod)

Doesn't cope with "from <...> import *" though; another reason not to
use that construct.

Cheers,
Michael



From gmcm at hypernet.com  Sat Nov  6 13:35:47 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sat, 6 Nov 1999 13:35:47 -0500
Subject: type object of an instance without instantiation?
In-Reply-To: <bgZU3.19722$Ec1.199882@news1.frmt1.sfba.home.com>
Message-ID: <1270209153-4344294@hypernet.com>

claudius writes:

> I have the following code (well, in summary):
> 
> class BigClass:
>  def __init__(self):
>   ...does some horrifyingly nasty operation...
> 
> def some_generic_fn(object):
>  if hasattr(object, '__class__') and object.__class__ ==
>  BigClass:
>   ...do some BigClass-specific thing...
> 
> Is there a way to create a type object for the instance of a
> BigClass object without actually creating an instance of one?

The type object of an instance (any instance) is the one and 
only InstanceType object. The type of object of a class (any 
class) is the one and only ClassType object.

I think you mean the BigClass class object. This exists as 
soon as the "class" statement is executed. When you create 
an instance of BigClass, you are "calling" the BigClass class 
object.

BTW, while doing lots of stuff in the constructor is often a 
valuable technique in other languages, it is much less so in 
Python. It makes things like pickling a lot harder (unless you 
provide default args and alternate ways to get the work done).
That doesn't mean it's *always* bad, just that it reduces your 
flexibility.

- Gordon



From kain at twilight.  Wed Nov  3 20:47:27 1999
From: kain at twilight. (Scott Barron)
Date: 4 Nov 1999 01:47:27 GMT
Subject: SWIG and function name question
References: <7vmc7e$8h$1@marcie.fissiontech.com> <m3ln8f9y9u.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <7vqojf$bgf$1@marcie.fissiontech.com>

Hi,

How about this:

Given these two files,

## foo.py
def __foo():
	print "Foo"

def bar():
	print "Bar"

## bar.py
import foo
class Foo:
	self.f = foo.__foo  # Create an alias for foo.__foo
	self.b = foo.bar    # Create an alias for foo.bar


>>> import bar
Traceback (innermost last):
  File "<stdin>", line 1, in ?
  File "bar.py", line 3, in ?
    class Foo:
  File "bar.py", line 4, in Foo
    self.f = foo.__foo
AttributeError: _Foo__foo

If I comment out self.f = foo.__foo all works fine.  What am I doing wrong
here?

Thanks,
Scott


In article <m3ln8f9y9u.fsf at atrus.jesus.cam.ac.uk>, Michael Hudson wrote:
>kain at twilight. (Scott Barron) writes:
>
>
>It's nothing to do with SWIG; try this for size:
>
>put this in a file called foo.py on your PYTHONPATH:
>
<snip>
>
>HTH,
>Michael
>
>O, look: http://www.python.org/doc/current/ref/import.html has all the
>gory details...



From mwh21 at cam.ac.uk  Fri Nov  5 10:15:58 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 05 Nov 1999 15:15:58 +0000
Subject: [OFFTOPIC] Maps and functions (was Re: Apply a function to each list member?)
References: <slrn824gee.1f7.bwinton@tor.dhs.org> <XFMail.991105110617.mikael@isy.liu.se> <19991105215504.A497@Ridcully.home> <14370.53219.667492.946075@gargle.gargle.HOWL> <27NwjKAK1uI4EwdJ@jessikat.demon.co.uk>
Message-ID: <m3eme59dg1.fsf@atrus.jesus.cam.ac.uk>

Robin Becker <robin at jessikat.demon.co.uk> writes:

> In article <14370.53219.667492.946075 at gargle.gargle.HOWL>, Charles G
> Waldman <cgw at fnal.gov> writes
[snip]
> >Hmm... what kind of "map" are you thinking of that could send 2 to
> >"both 3 and 4"?  I think such a thing is not a mapping, at least not
> >according to any definition of "map" that I've ever seen.  If you
> >look, for instance, in any basic topology text, you will see that
> >"map" and "function" mean exactly the same thing, when used as nouns.
> >"Map" also has a sense as a transitive verb, which means "to apply a
> >function to", which is where the Python/functional programming usage
> >comes from.
> >
> >
> >
> I define f as a thing which takes integer arguments and outputs sets of
> integers
> 
> presumably I can have such an f that
> 
>         f(2)={3,4}
> 
> to my simple engineer's eyes f could belong to the functions mapping Z--
> >2^Z
> 
> I assume we're not talking that way if the f above isn't a function.

f is a function, as you say from Z to 2^Z. A difference between that
and a function from Z to Z, is that this f throws away a lot of the
algebraic structure of Z (i.e. being a uniquely generated commutative
ring), and so is in many ways less interesting than a function that
preserves this structure.

Regards,
Michael



From phd at phd.russ.ru  Fri Nov 19 03:50:23 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Fri, 19 Nov 1999 08:50:23 +0000 (GMT)
Subject: urllib.getsize()
In-Reply-To: <14388.29839.294164.707388@dolphin.mojam.com>
Message-ID: <Pine.LNX.4.21.9911190849410.12519-100000@fep132.fep.ru>

On Thu, 18 Nov 1999, Skip Montanaro wrote:
>     Gerrit> I've written a function to get the size of a HTTP/FTP url, is it
>     Gerrit> good enough to be put in urllib.py?
> 
> My only worry is that HEAD doesn't always work (for whatever reason having
> to do more with servers than clients).  Would it be better to wrap things in
> a try/except block?  Anyone have a URL that fails with HEAD?

   What is worse, not every server returns Content-Length. I wrote many
CGIs that didn't :)

Oleg.
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From kuncej.CENTRAL2.MDCEMAIL at mail.conservation.state.mo.us  Fri Nov 19 12:50:24 1999
From: kuncej.CENTRAL2.MDCEMAIL at mail.conservation.state.mo.us (Jeffrey Kunce)
Date: Fri, 19 Nov 1999 11:50:24 -0600
Subject: Batch window
Message-ID: <s83539bf.035@mail.conservation.state.mo.us>

>Is it posible to automaticaly minimize the python window, when starting
>a GUI application on Windows ?

As others have answered, pythonw.exe will run a python program
with *no* python window.

If you actually want a minimized python window to be present,
use something like:

    START /MIN python.exe myprog.py

Sometimes the python window is useful for debugging, or
as a log window.

  --Jeff




From Markus_Strickler at everest-design.de  Thu Nov 11 11:56:07 1999
From: Markus_Strickler at everest-design.de (Markus Strickler)
Date: Thu, 11 Nov 1999 17:56:07 +0100
Subject: AW: Python 1.5.2 on NeXTStep
In-Reply-To: <199911111629.AA23443@ch1d2833nwk>
Message-ID: <002901bf2c65$a5edb000$2d131bc3@everestdesign.com>

Hi!
Thanks for your reply.
I would appreciate any help I can get. So if you could send me your modfied
sources or whatever you have that would be great.
Thanks again,

Markus Strickler

> -----Urspr?ngliche Nachricht-----
> Von: Ted Horst [mailto:Ted.Horst at wdr.com]
> Gesendet am: Donnerstag, 11. November 1999 17:30
> An: Markus Strickler
> Betreff: Re: Python 1.5.2 on NeXTStep
>
>
> I have compiled 1.5.2 on our systems which are 4.2 with 3.3 developer (I
> think this is the same as compiling on 3.3).  Unfortunately, I
> had to change
> some of the source, and I never got around to making good diffs.  I can
> send you what I have if you want.
>
> On Thu, 11 Nov 1999, "Markus Strickler" wrote:
> > Hi!
> > Has anyone managed to compile Python 1.5.2 on NeXTStep 3.3? I'm
> currently
> > running an older Python version and would like to upgrade. However 1.5.2
> > won't compile. Can anyone give me start what config parameters I need.
> >
> > Thanks,
> >
> > Markus
> >
> >
> > --
> > <http://www.python.org/mailman/listinfo/python-list
> -----------------------------
> Ted Horst <Ted.Horst at wdr.com>
> "the day serious reasoning about programs becomes widespread is the day
> most programmers abandon their craft to resume drug trafficking <wink>."
> --Tim Peters on comp.lang.python
>




From gmcm at hypernet.com  Mon Nov 15 11:33:49 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Mon, 15 Nov 1999 11:33:49 -0500
Subject: win32api... another question:  GetEnvironmentVariable("PATH")
In-Reply-To: <W8WX3.67146$Rx2.793309@news1.rdc1.bc.home.com>
Message-ID: <1269438868-3874906@hypernet.com>

Kevin Cazabon wrote:

> Interesting... it doesn't work on Win98 for me (The exact statement
> below...).
> 
> Anyone have suggestions for Win95/98?
> 
> 
> Kiss, Arpad <AKiss at GEOMETRIA.hu> wrote in message
> news:01BF2F61.1D2B8DB0 at exchange1.geometria.hu...
> > Hi,
> >
> > Kevin Cazabon <kcazabon at home.com> wrote in message
> > news:<fcMX3.65012$Rx2.773564 at news1.rdc1.bc.home.com>...
> > > I can't seem to get the win32api.GetEnvironmentVariable() function to
> > work
> > > properly... am I not using it correctly by the following?
> > >
> > > path = win32api.GetEnvironmentVariable("PATH")
> >
> > It works for me on a Windows NT Workstation (sp5).

First, why not use os.environ['PATH']?

Second, while I'm on NT and 'PATH' and 'path' return the same 
answer (both the os and the win32api way), I had a bug report 
from someone on Win9x and the solution turned out to be:
 try:
   value = os.environ[key]
 except KeyError:
   value = os.environ[string.lower(key)]



- Gordon



From gmcm at hypernet.com  Mon Nov  1 07:24:47 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Mon, 1 Nov 1999 07:24:47 -0500
Subject: os.popen() and (IDLE, PythonWin) ?
In-Reply-To: <381D63CA.9073803D@lemburg.com>
Message-ID: <1270664157-24361801@hypernet.com>

Marc-Andre wrote:
> "M.-A. Lemburg" wrote:
> > 
> > [os.popen('ver') fails under IDLE and PythonWin, yet works
> > under "normal"
> >  Python]
> 
> Ok, I found the FAQ entry on this:
> 
> 7.13. Why doesn't os.popen() work in PythonWin on NT?
[snip]
> Under PythonWin this is no problem (although I wish os.popen
> would automatically choose the different approach for me in the
> above setup). Yet under IDLE I have a problem: win32 doesn't have
> to be installed... is there some other workaround ?

tempfile and redirection :-(.

- Gordon



From b2blink at hotmail.com  Thu Nov 18 08:51:33 1999
From: b2blink at hotmail.com (Ulf Engstrøm)
Date: Thu, 18 Nov 1999 08:51:33 CET
Subject: Python in Windows CE?
Message-ID: <19991118075134.9052.qmail@hotmail.com>

>   Has anybody used Python for development under Windows CE?
>--
>rshory at nrel.colostate.edu

Well, not really using it yet, but I've been trying to gather some info 
about it, cause by next week I'll start doing some Python on a small machine 
running CE. I'd also be interested in all info that anyone might have and 
I'm glad to answer any questions about my development once I get it going.

Did you Rick have any particular questions about running under CE or was it 
just a general question?

Best regards
Ulf Engstr?m

http://www.b2b-link.com
ulf.engstrom at b2b-link.com

______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com



From fdrake at acm.org  Wed Nov 17 13:36:34 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Wed, 17 Nov 1999 13:36:34 -0500 (EST)
Subject: [Reminder] IPC8 Call For Demos and Posters
Message-ID: <14386.62898.780630.727260@weyr.cnri.reston.va.us>

  The deadline for demo & poster proposals for the next Python
conference is approaching; please note these important dates:

    Nov. 30 : Deadline for demos and posters.
    Dec. 15 : Notification of acceptance for demos and posters.

  More information can be found at:

    http://www.python.org/workshops/2000-01/demos.html

or send email to me at <fdrake at acm.org>.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From andy at robanal.demon.co.uk  Thu Nov  4 16:23:05 1999
From: andy at robanal.demon.co.uk (Andy Robinson)
Date: Thu, 04 Nov 1999 21:23:05 GMT
Subject: Assorted Questions regarding the future of Python...
References: <004201bf257b$0b0c64f0$0501a8c0@bobcat>
Message-ID: <3822f253.11537059@news.demon.co.uk>

"Mark Hammond" <mhammond at skippinet.com.au> wrote:

>> From: python-list-admin at python.org
>> [mailto:python-list-admin at python.org]On Behalf Of Rob Hodges
>>
>> Unless something has changed dramatically since I last used windoze,
>> VB is not compiled (in the sense of native machine code) any more
>than
>> a frozen python program; rather, VB "executables" have a small
>header
>> of native code that launches the vbrunxxx.dll (interpreter) to run
>the
>> bytecode that makes up the rest of the file.  Which is more or less
>
>Actually, things have changed.  The latest version of VB has an option
>to compile to native machine code.  This feature was created more from
>a "check-box marketting" POV than anything else (ie, Delphi did it, so
>VB decided it must too!)
>
>Mark.
When VB's native compilation came out, some Delphi people were quick
to time it and discovered that Delphi was still about 10x faster for
typed math and indexing operations, although the benefits disappeared
when dealing with variants or (as you do most of the time) COM APIs to
other things.  If they read things right, it does compile, but not
desperately intelligently.

I haven't done any benchmarking myself, but when I installed VB6 I
didn't exactly leap off my seat at the speed increase.  

Andy



From bernhard at alpha1.csd.uwm.edu  Tue Nov 30 13:15:42 1999
From: bernhard at alpha1.csd.uwm.edu (Bernhard Reiter)
Date: 30 Nov 1999 18:15:42 GMT
Subject: XPython (X11)
References: <383E5350.1E07536A@xrce.xerox.com> <slrn83tj53.k7h.bernhard@alpha1.csd.uwm.edu> <81ms2n$126$1@essle.valdyas.org> <slrn840dt7.mov.bernhard@alpha1.csd.uwm.edu> <m3hfi64oyh.fsf@phaedrus.zen.xom> <slrn845im9.178.bernhard@alpha1.csd.uwm.edu> <81vsg2$g8$2@essle.valdyas.org>
Message-ID: <slrn84852e.cn.bernhard@alpha1.csd.uwm.edu>

On 30 Nov 1999 06:57:06 GMT, Boudewijn Rempt <boud at rempt.xs4all.nl> wrote:
>Bernhard Reiter <bernhard at alpha1.csd.uwm.edu> wrote:

>> True. But I want that over an integrated API and GNOME and KDE
>> don't offer that yet. And they also don't offer it cross platform.
>> wxWindows at least tries to get there.
>
>I'm curious as to what you mean by an 'integrated API' - just
>curiosity.

Bad wording on my part, I guess. :)
Maybe I should call it internally aimed API.
Means: Yes, you can use it from outsite, but the abstraction in this
case is wrong.

Like: Why do I need a HTML widget if I can just delegate the task to
any HTML browser. The GNOME people would never delegate a task to the
KDE HTML widget. But they could have an API to delegate it to any
HTML browser and then let the user assemble it. If the user has
KDE on the machine, he might as well use the KDE HTML browser.

This is why the KDE and GNOME APIs are targeted within their own
structure, but not to seperate tasks and nurture compatibility.

>>>And yeah, last I heard you had to buy the Troll Kit for PyQt under
>>>Win32.  
>> Dropping platform compatibility, that a huge point.

>Oh well, platform compatibility that costs money is still
>platform compatibility.
No. Proprietory software is not a real option. 
	Bernhard
-- 
Research Assistant, Geog Dept UM-Milwaukee, USA.  (www.uwm.edu/~bernhard)
Free Software Projects and Consulting 		         (intevation.net)  
Association for a Free Informational Infrastructure            (ffii.org)



From mwh21 at cam.ac.uk  Sat Nov  6 12:50:23 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 06 Nov 1999 17:50:23 +0000
Subject: making the Python case to my dept.
References: <9911051740130D.02530@william.jinsa.org> <3823FC43.CF7FBD71@python.net>
Message-ID: <m3puxno6g0.fsf@atrus.jesus.cam.ac.uk>

A couple of little points...

"Thomas A. Bryan" <tbryan at python.net> writes:

[biggo snippo]
> code in another file.  Then, if they can't remember the methods for 
> a list, they can do something like this
> 
> >>> L = []
> >>> dir(L)
> ['append', 'count', 'extend', 'index', 'insert', 'pop', 'remove',
> 'reverse', 'sort']

``import rlcompleter''!!

It makes life *so* much easier, and is the reason I still use the
basic interpreter rather than idle.

Hmm, implementing completion in idle? It has ``expand word'' but
that's not what I want. Would this be easy? Pointers appreciated.

[snippo two]
> Do math majors or engineers learn their critical job skill's in their 
> first classes?  

What are mathematicians' critical job skills? I'm curious, being one,
and not having being taught anything resembling a `critical job skill'
yet, and happy about it.

Regards,
Michael



From ionel at psy.uva.nl  Fri Nov 19 10:09:28 1999
From: ionel at psy.uva.nl (Ionel Simionescu)
Date: Fri, 19 Nov 1999 16:09:28 +0100
Subject: Comparing Python to (free) Math languages/environments
References: <19991118014716.13748.00000334@ng-bj1.aol.com> <810rev$jsm$1@nnrp1.deja.com> <874sej7vmg.fsf@ifm.uni-kiel.de> <8134cl$ab1$1@nnrp1.deja.com>
Message-ID: <813p47$jlh@mail.psy.uva.nl>

----- Original Message -----
From: <siegfried.gonzi at kfunigraz.ac.at>
[...
Siegfried explaying how easy one can get started with Matlab and the likes
to solve scientific problems.
...]


Dear Sigfried,

You are right - in the context of dealing with a few small, "code and throw
away" problems.

When you shall develop lots of interrelated pieces of code, and when you
*and others* will use that code to run various, critical experiments, you
will need:
- database support to manage your data,
- GUI's to let others use your programs before reading a ton of docs,
- new libraries to cover problems that your specialized language of choice
did not think yet about,
and
- utilities to help with documenting and debugging your code.

Shortly put, you will need a real programming language.

Thus, while I respect your choice and find it right, you may reconsider it
if you are going to develop programs for more than one year.

Note that I do not push you to Python, even if it is my choice, - feel free
to choose whatever language you like better. But invest your time wisely -
start with proper tools if you intend to build something bigger than
programs for a diploma.


ionel







From mlh at vier.idi.ntnu.no  Fri Nov 26 08:53:13 1999
From: mlh at vier.idi.ntnu.no (Magnus L. Hetland)
Date: 26 Nov 1999 14:53:13 +0100
Subject: CRC-module
References: <38396766.B189D15A@bibsyst.no> <383C1089.1D0F7C18@bibsyst.no> <slrn83p9o2.5pm.davecook@rama.escnd1.sdca.home.com>
Message-ID: <y0jr9hdxsty.fsf@vier.idi.ntnu.no>

davecook at home.com (David M. Cook) writes:

> On Wed, 24 Nov 1999 17:21:29 +0100, Thomas Weholt <thomas at bibsyst.no> wrote:
> 
> >Ok, so I`ve looked into zlib.crc32 and zlib.adler32. They seem easy
> >enough to use, but I thought crc-codes had characters and numbers in
> >them, 
> 
> Could the characters be a-f?  Maybe apply hex() to the number?

sha.sha("Hello, world!").hexdigest()

> 
> Dave Cook

--

  Magnus          Echelon jamming noise:
  Lie             FBI CIA NSA Handgun Assault Bomb Drug Terrorism
  Hetland         Special Forces Delta Force AK47 Hillary Clinton 



From darcy at vex.net  Thu Nov 11 12:53:44 1999
From: darcy at vex.net (D'Arcy J.M. Cain)
Date: 11 Nov 1999 17:53:44 GMT
Subject: ASCII delimited files
References: <942231502.1914033961@news.interbusiness.it> <80ekhb$1qrh$1@hub.org> <80et0b$2df8$1@hub.org>
Message-ID: <80evr8$2lc7$1@hub.org>

D'Arcy J.M. Cain <darcy at vex.net> wrote:
> And here it is.  Note that Christian Tismer wrote the original (at least

Skip it.  There's a bug.  If a line ends with a quoted entry it gets
attached to the next one.  I'm tracking it down now and will post the
corrected one.

I have already added the test to __main__.  :-)

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



From gmcm at hypernet.com  Tue Nov 16 12:12:05 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 16 Nov 1999 12:12:05 -0500
Subject: function override???
In-Reply-To: <80s0sd$1g4k$1@rtpnews.raleigh.ibm.com>
Message-ID: <1269350172-9209840@hypernet.com>

Arinte writes:

> I am defining a python callback for my embedded app, but I want
> the script writer to override my function.  So, basically what I
> want is a function prototype.  Ex. here is what I am thinking.
> 
> ;---fprot.py
> def overrideme(self, str, someint)
> 
> ;--fimp.py
> import fprot   #I know this ain't right.
>                           # I don't know how to import from other
>                           python
> scripts yet.  Enlighten me?
> 
> def overrideme(self, str, someint)
>     do_some_junk()

Is that a function or a method?

If it's a function, I would just document the signature. If it's a 
method, I would expect the user to do:

import fprot

class MyThing(fprot.Thing):
  def overrideme(self, strng, someint):
     .....

In either case, when I'm implementing a callback, my first step 
is to say something like:

def callback(*args):
  print repr(args)

and see what turns up.

- Gordon



From gerrit.holl at pobox.com  Sun Nov 28 14:04:17 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Sun, 28 Nov 1999 20:04:17 +0100
Subject: Parsing text
In-Reply-To: <g1e04.1$O8.108@news.inreach.com>; from trekkan@yahoo.com on Sun, Nov 28, 1999 at 11:01:30AM -0800
References: <g1e04.1$O8.108@news.inreach.com>
Message-ID: <19991128200417.A2508@stopcontact.palga.uucp>

Troy wrote:
> As you will be able to tell, I am new to Python so please forgive any newbie
> things I might do. =)
> 

[string parsing questions]

Troy,

read the documentation on the string module. It's likely to be on your HD,
but it's also on www.python.org/doc

Success!

regards,
Gerrit.

-- 
"The move was on to 'Free the Lizard'"

  -- Jim Hamerly and Tom Paquin (Open Sources, 1999 O'Reilly and Associates)
  8:03pm  up  2:28, 13 users,  load average: 1.52, 1.58, 1.71



From ivanlan at callware.com  Tue Nov 23 09:04:36 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Tue, 23 Nov 1999 07:04:36 -0700
Subject: Python books
References: <199911230613.HAA21179@dionysus.fw.cuci.nl>
Message-ID: <383A9EF4.5D1E8986@callware.com>

Hi All--

Hans Nowak wrote:
> 
> A quick search at Proxis (online bookstore, http://owas.proxis.be)
> for Python yields, among others, the following books:
> 
> John Grayson: Python & Tkinter programming
> Ivan van Laningham: Python in 24 hours
> ?: Python Annotated Archives
> 
> It claims that all these are published already. Does anyone know if
> this is really the case? Proxis has been known to put "published"
> books in their list long before they actually hit the shelves.
> 

My book wasn't originally scheduled till December.  Publication date now
looks like February.

John Grayson's book is finished, but I suspect that publication will
also be February.

If "Python Annotated Archives" is /F's book, it's already out as an
eMatter book.  Visit www.python.org for details on how to get it. 
Search Deja News for important information on using this style of
book--there was a thread on here about it.

Take care,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours
----------------------------------------------



From foxes at theriver.com  Fri Nov 12 11:51:35 1999
From: foxes at theriver.com (Andy Fox)
Date: Fri, 12 Nov 1999 09:51:35 -0700
Subject: '==' vs. 'is' behavior
References: <382C2C6A.64A8F2C2@theriver.com>
Message-ID: <382C4597.331DE73A@theriver.com>

Hi,

Thanks for setting me straight.  Like many things, it makes sense when
the underlying reason is understood.

I'm not used to using languages where the source is available, so  I had
not considered this option.  I'm downloading the source now, and will
try not to ask any more silly questions ;-)  

-- 
Andy Fox



From JamesL at Lugoj.Com  Thu Nov 25 00:25:09 1999
From: JamesL at Lugoj.Com (James Logajan)
Date: Wed, 24 Nov 1999 21:25:09 -0800
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <867lj85g4c.fsf@g.local>
Message-ID: <383CC835.B585EBDD@Lugoj.Com>

Gareth McCaughan wrote:
>   - if/then/else works on *statements* (strictly, suites of
>     statements) rather than on *expressions*, so I can't
>     say
>       print "You scored %s point%s" % (score, if score==1 then "" else "s")
>     or anything like that. (Or, as some people prefer to put it,
>     Python has no ? : operator.)

Thanks to Python's dynamic typing you can easily add the C "?:" expression
to the language by defining a single function that will work with all types:

def ifexp(cond, trueVal, falseVal):
   if cond:
      return trueVal
   else:
      return falseVal

print "You scored %s point%s" % (score, ifexp(score == 1, "", "s"))

Note that you'd have to define a whole set of these functions in a
statically typed language, one for each type of the Vals.

>   - Memory management is by refcounting instead of GC. This
>     has its good points (it means that one can define more
>     precisely when e.g. finalisation happens), but it does
>     mean that if you build structures with cycles in the
>     reference graph then they won't get GCed when you want
>     them to. This is a real nuisance, for instance, when
>     you're exploring ideas or debugging: suppose you have
>     a piece of code that builds a big data structure out of
>     class instances; then every time you make a small change
>     and run it again (you have to run it again, because
>     existing instances belong to the old versions of the
>     classes) you can be leaking a lot of memory.

Sorry, but I think the price paid for "true" GC is too high for the small
number of programs that benefit from this. Perhaps one should learn to clean
up after oneself when cycles are expected.



From tbryan at python.net  Wed Nov 10 21:20:49 1999
From: tbryan at python.net (Thomas A. Bryan)
Date: Wed, 10 Nov 1999 21:20:49 -0500
Subject: ASCII delimited files
References: <942231502.1914033961@news.interbusiness.it> <382A240B.B247AD6B@python.net>
Message-ID: <382A2801.FDE3CA09@python.net>

"Thomas A. Bryan" wrote:

> class DelimParserField:
>     def __init__(self, name):
>         self.name = name
>     def convert(self,value):
>         return value
>     def verify(self,value):
>         pass
> 

One important thing that I didn't point out.  The "name" member 
is the column's name.  (Remember that I was thinking of 
delimited files for import into a database when I wrote these 
classes.)

Thus, a more sensible application might have something like
employeeListParser = DelimFldParser((EnumField('Department',('HR','IT')),
                                     DelimParserField('Name'),
                                     NumericRngField('Salary',0,500000)))

If anyone uses this stuff, I'd be interested in hearing about it.

---Tom



From viruswall at bam.com  Thu Nov 25 17:00:04 1999
From: viruswall at bam.com (viruswall at bam.com)
Date: Thu, 25 Nov 1999 17:00:04 -0500 (Eastern Standard Time)
Subject: InterScan NT Alert
Message-ID: <199911252205.RAA13525@python.org>

Receiver, Bell Atlantic Mobile has detected virus(es) in the e-mail attachment.

Date:  	Thu, 25 Nov 1999 17:00:04 -0500 (Eastern Standard Time)
Method:	Mail
From:  	<botnivick at hotmail.com>
To:    	<python-list at python.org>
File:  	Happy99.exe
Action:	clean failed - deleted
Virus: 	TROJ_SKA 



From robin at jessikat.demon.co.uk  Sun Nov 28 20:42:49 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Mon, 29 Nov 1999 01:42:49 +0000
Subject: http://www.python.org/ blank ?
References: <HJYc$DA+GcQ4EwYP@jessikat.demon.co.uk>
Message-ID: <o4hHjFAZodQ4EwPX@jessikat.demon.co.uk>

In article <HJYc$DA+GcQ4EwYP at jessikat.demon.co.uk>, Robin Becker
<robin at jessikat.demon.co.uk> writes
>for some reason I see http://www.python.org/ as a blank page. Any ideas?
actually I see it as blank (pure white) with NS 4.7 win95. It seems ok
with IE 5. I can see other pages with NS ok eg
http://www.python.org/search/

Weird; it just came back ok. My apologies.
-- 
Robin Becker



From mwh21 at cam.ac.uk  Sun Nov 28 14:34:53 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 28 Nov 1999 19:34:53 +0000
Subject: Parsing text
References: <g1e04.1$O8.108@news.inreach.com>
Message-ID: <m3k8n2s942.fsf@atrus.jesus.cam.ac.uk>

"Troy" <trekkan at yahoo.com> writes:

> As you will be able to tell, I am new to Python so please forgive any newbie
> things I might do. =)

Welcome!

> What I would like to do is take a variable and parse out each of the
> pieces.  Most commonly delimited by spaces.  Example text:
> 
> This is a test
> 
> I would like to be able to pull out any of "This", "is", "a", "test"
> and be able to drop any element into a variable that I wanted.

string.split?

I.e. 

Python 1.5.2+ (#12, Nov 19 1999, 15:00:31)  [GCC 2.95.2 19991024 (release)] on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import string
>>> string.split("This is a test")
['This', 'is', 'a', 'test']

> I hope this explains it enough.  Thanks for any help you may be able
> to provide.
> 
> One other thing, is there a way to tell if a substring exists in a
> string?  in a basic sort of way
> 
> if 'abc' isin '123abc456'

string.find?

>>> string.find("1122abc33","abc")
4
>>> string.find("1122abc33","f")
-1

> Something to that effect anyway.

In general

http://www.python.org/doc/current/lib/module-string.html

and maybe

http://www.python.org/doc/current/tut/tut.html

?

> Thanks for your time!

No problem!

Cheers,
Michael



From thomas at xs4all.nl  Sat Nov 13 20:14:55 1999
From: thomas at xs4all.nl (Thomas Wouters)
Date: Sun, 14 Nov 1999 02:14:55 +0100
Subject: .dll and .exe file version informaton
In-Reply-To: <000101bf2b3c$a7d43e40$d32d153f@tim>; from tim_one@email.msn.com on Wed, Nov 10, 1999 at 12:30:10AM -0500
References: <806tck$5o9$1@nnrp1.deja.com> <000101bf2b3c$a7d43e40$d32d153f@tim>
Message-ID: <19991114021455.A23596@xs4all.nl>

On Wed, Nov 10, 1999 at 12:30:10AM -0500, Tim Peters wrote:

> [emuller at painewebber.com]
> > Can anyone give me an idea how to quikly get the version info on .dll
> > and .exe files under windows?

> In Explorer, right-click on the file, select Properties from the context
> menu, and click the Version tab.

> go-ahead-just-*try*-to-write-a-program-to-do-that<wink>-ly y'rs  - tim

Chuckle. This is not as funny as it seems. We actually do that. That is, the
company I work for has a windows 'Internet Kit', which installs all kinds of
programs for you, configures dialup networking, etc -- by doing exactly what
you describe. It only has a few drawbacks... it requires the user to sit
back and not touch the keyboard (but then, it's intended for people who do
not want to hassle with their PC's setup) and it's written in VB ;)

Of course, the Kit uses dirty tricks to avoid OCR'ing an application window
to find out what it says ;)

Typical-windows-app-it-only-goes-wrong-with-the-clueless'ly y'rs,
-- 
Thomas Wouters <thomas at xs4all.net>

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



From endgamer71 at hotmail.com  Mon Nov  8 08:34:11 1999
From: endgamer71 at hotmail.com (Endgamer)
Date: Mon, 08 Nov 1999 13:34:11 GMT
Subject: Newbie question: Input
Message-ID: <3826b693.3888197@212.225.36.99>

Okay, I know it's a damn stupid question, but I've been through the
books, the texts, the tutorial, the language reference and no luck...

Whats the syntax to prompt for user input?


Endgamer
Spider Warrior

(Endgamer71 at spamtrap.hotmail.com)
#####################################
Not a quote, just an on-the-spot observation:
You know you're a geek when your PC costs exactly
15 times as much as your stereo. And gives better
sound quality.<sigh>
                              Endgamer
#####################################



From thantos at chancel.org  Thu Nov 25 23:43:21 1999
From: thantos at chancel.org (Alexander Williams)
Date: Fri, 26 Nov 1999 04:43:21 GMT
Subject: The First Annual Obfuscated Python Contest
References: <mfbZ3.71$AZ6.1807@dfiatx1-snr1.gtei.net> <001d01bf328d$3ac77170$f29b12c2@secret.pythonware.com>  <slrn83s2vv.1s0t.cjc26@tankgrrl.bridget.mindriot.net>
Message-ID: <slrn83s46n.cvu.thantos@chancel.org>

On 26 Nov 1999 04:27:46 GMT, Cliff Crawford <cjc26 at nospam.cornell.edu> wrote:
>next-they'll-want-obfuscated-Haskell-ly yr's, Cliff

Never.  Redundant.  :)

-- 
Alexander Williams (thantos at gw.total-web.net)
"In the end ... Oblivion Always Wins."



From gmcm at hypernet.com  Sun Nov 21 13:15:45 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sun, 21 Nov 1999 13:15:45 -0500
Subject: Order of keyword arguments
In-Reply-To: <383822C3.DD0CE177@corrada.com>
Message-ID: <1268914352-35423797@hypernet.com>

Andres Corrada wrote:

> Jeff wrote:
> 
> > ....................................................... but I
> > don't see the advantage to 'knowing' the order that the
> > keywords were used by the calling function.
> > 
> 
> I think there would be an advantage to having such a data
> structure as an ordered dictionary. My "American Heritage
> Dictionary of the English Language" is such a data structure.

Ah, but that would be "alphabetical order", not "entry order", 
and not much use to anybody if it were.

Python data structures that would suit your apparent desires:
 [ (optionname, optionvalue), ... ]

 [ optionname, ...], {optionname: optionvalue, ...}

You could subclass UserList to provide the intended 
semantics, and then force your poor, hapless user to pass in 
exactly one arg (an instance of your UserList subclass).

Or you could ponder the meaning of cheese. 

With approximately equal utility.

- Gordon



From thor at localhost.localdomain  Sun Nov 28 13:45:15 1999
From: thor at localhost.localdomain (Manuel Gutierrez Algaba)
Date: 28 Nov 1999 18:45:15 GMT
Subject: [INSIST] Python ring
Message-ID: <slrn8431p9.md.thor@localhost.localdomain>

This is a reminder that there's a ring for python sites,
available for *ANYONE* that has a web page about python,
the main web page is at:
http://www.ctv.es/USERS/irmina/pythonring.html

This is supported by webring.com

Please, consider join the ring, I think it's a good think
for the community

Thanks
-- 




From gmcm at hypernet.com  Sat Nov  6 13:35:47 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sat, 6 Nov 1999 13:35:47 -0500
Subject: Job skills [was Re: making the Python case to my dept.]
In-Reply-To: <m3puxno6g0.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <1270209151-4344374@hypernet.com>

Michael Hudson asks:

> "Thomas A. Bryan" <tbryan at python.net> writes:
> 
> > Do math majors or engineers learn their critical job skill's in
> > their first classes?  
> 
> What are mathematicians' critical job skills? I'm curious, being
> one, and not having being taught anything resembling a `critical
> job skill' yet, and happy about it.

The ability to work long and hard on entirely meaningless 
problems.

white-space-anyone?-ly y'rs

- Gordon



From ScherBi at BAM.com  Tue Nov 16 11:46:35 1999
From: ScherBi at BAM.com (ScherBi at BAM.com)
Date: Tue, 16 Nov 1999 11:46:35 -0500
Subject: Anyway to use file masks with os.listdir()?
Message-ID: <3F8BA119B48AD11197470008C7A493CB98B67A@HQOBGEX1>

use glob.

eg:

DIR = "/home/spam"
files = glob.glob(DIR + "/*.jpg")

will get you a list of all the .jpg files in /home/spam


Bill Scherer

> -----Original Message-----
> From: Aaron J Reichow [mailto:reic0024 at ub.d.umn.edu]
> Sent: Tuesday, November 16, 1999 10:29 AM
> To: python-list at python.org
> Subject: Anyway to use file masks with os.listdir()?
> 
> 
> 
> Is there any way to pass a filemask to os.listdir()?  For 
> example, to list
> only files with the extension .jpg.  Any help would be appreciated! :)
> 
> Aaron
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 



From gmcm at hypernet.com  Fri Nov 19 09:14:29 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 19 Nov 1999 09:14:29 -0500
Subject: Beta 3e of the Win32 Installer
In-Reply-To: <1269243486-15626894@hypernet.com>
Message-ID: <1269101629-24159362@hypernet.com>

I wrote:
> Win32 Installer users:
> 
> Beta 3e is here (the *last* beta before release 3, got that,
> Burt? <wink>).

And here's the current bug list:
 - the run from CDROM stuff doesn't work. I have a fix which I'll 
email if you need it now
 - the change-the-icon stuff apparently only works on NT
 - it also will fail unless you have Mark's Win32 stuff installed 
(this one I have a fix for).
 
I'm keeping a buglist on the download page, so I won't have to 
bother the group with these messages.
 
> The new beta features:
>  - run from CDROM (if R/O, uses temp directory)
>  - put your own icon in the .exe (thanks to Thomas Heller!)
>  - bug fixes
>  - more docs
>  - examples of how to put Python on a floppy.
> 
> In fact, one of the configs will put Python, Tcl/Tk and IDLE on a
> floppy. Use it interactively, or pass it a script! Just like real
> Python.
> 
> Check out http://starship.python.net/crew/gmcm/install.html
> 
> A more formal announcement will follow when it goes release.
> 
> - Gordon
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list



- Gordon



From tim_one at email.msn.com  Fri Nov 26 22:48:38 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Fri, 26 Nov 1999 22:48:38 -0500
Subject: filter function question - list processing in Python
In-Reply-To: <81f18b$hor$1@nnrp1.deja.com>
Message-ID: <000701bf388a$4a0ef600$742d153f@tim>

[lobozc at my-deja.com]
> ...
> In general, I note that Python has some standard functions operating
> on lists but they are curiously limited.

If you're talking about filter, map and reduce, they were contributed
additions to the language, and Guido has said that accepting the patch was
probably a mistake.

> It is a pity, since these are the functions which reduce line count
> significantly - as well as make programming safer. I wonder what are
> the chances of repeating with lists what Python did with regular
> expressions - copying the functionality from a well tested and popular
> product?

In light of the above, there's no chance that more functions of this kind
will be added (that doesn't stop you from writing your own -- it's easy).

OTOH, Haskell-like list comprehensions got a good reception last time they
were discussed, and Greg Ewing has implemented a version of them.  I hope
they make it into the base language.  A good part of their attraction to
Guido was that they provide a Pythonic way to get the map and filter
functions per se *out* of the language.

Note also that the NumPy extension adds list-processing of the "array
syntax" flavor (flexible N-dimensional slicing, data-parallel SIMD, and
assorted reduction and reshaping functions).

map-without-currying-is-like-thanksgiving-without-turkey-ly y'rs  - tim





From wheinemanNOwhSPAM at uconect.net.invalid  Tue Nov 16 10:20:57 1999
From: wheinemanNOwhSPAM at uconect.net.invalid (wheineman)
Date: Tue, 16 Nov 1999 07:20:57 -0800
Subject: Need help dealing with spaces in NT paths.
References: <80pr76$jq5$1@nnrp1.deja.com> <11f733ec.96958819@usw-ex0101-007.remarq.com> <uogcu93dt.fsf@cs.uu.nl>
Message-ID: <0a0133f8.1afd50af@usw-ex0102-014.remarq.com>

In article <uogcu93dt.fsf at cs.uu.nl>, piet at cs.uu.nl wrote:
> >>>>> wheineman <wheinemanNOwhSPAM at uconect.net.invalid> (W) writes:
> W> Put quotes around the complete path. For example:
> W> myExec = "C:\\Program Files\\Trash Can\\compactor.exe"
> You probably mean:
> myExec = "\"C:\\Program Files\\Trash Can\\compactor.exe\""
> or easier:
> myExec = r'"C:\Program Files\Trash Can\compactor.exe"'
> --
> Piet van Oostrum <piet at cs.uu.nl>
> URL: http://www.cs.uu.nl/~piet [PGP]
> Private email: Piet.van.Oostrum at gironet.nl

What he said. Sorry.


* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!




From jeremy at cnri.reston.va.us  Fri Nov 12 15:32:03 1999
From: jeremy at cnri.reston.va.us (Jeremy Hylton)
Date: Fri, 12 Nov 1999 15:32:03 -0500 (EST)
Subject: method overriding trick
In-Reply-To: <Pine.WNT.4.04.9911110920120.207-100000@rigoletto.ski.org>
References: <14378.62498.649973.150075@bitdiddle.cnri.reston.va.us>
	<Pine.WNT.4.04.9911110920120.207-100000@rigoletto.ski.org>
Message-ID: <14380.31043.118853.946057@goon.cnri.reston.va.us>

I suggested a renaming approach for managing method overriding.

class Base:
    def aMethod(self, arg):
        return arg * 2

class Derived1(Base):
    super_aMethod = Base.aMethod

    def aMethod(self, arg):
        return self.super_aMethod(arg) + 1

David suggested implementing the same idea via a getattr hook that
eliminated the need for explicit naming.

import new
class Mixin:
    def __getattr__(self, k):
	if k[:len('super_')] == 'super_':
	    for b in self.__class__.__bases__:
		m = getattr(b, k[len('super_'):], None)
		if m: 
		    return new.instancemethod(m, self, self.__class__)
	    else:
		raise AttributeError, k
	else:
	    raise AttributeError, k

class Base(Mixin):
    def f(self): 
	print 'Base.f'
    
class Sub(Base):
    def f(self):
	print 'Sub.f'
	self.super_f()

I think this approach might work, but there are some details you've
skated over that would need to be worked out.  The main problem is
dealing with an inheritance hierarchy where there are more levels than
two.

class SubSub(Sub):
    def f(self):
        print "SubSub.f"
        self.super_f()

I you call method f on a SubSub object, you'll dump core.  When the
getattr hook is invoked, it will create a method named super_f for the
object.  Because it is first called in SubSub, the getattr hook will
bind super_f to Sub.f.  But Sub.f also references the name super_f, so
it will call itself.

Actually, you'll hit this problem any time you call f in a class
derived from one that implements a method named f.  Even if SubSub
didn't override f, the search for an f would find the most derived
implementation first. 

I think you can solve the problem with private variables.
self.__super_f will get mangled at compile time and you can do a more
complicated search for attribute names that look like they've been
mangled. 

You would also need to do a proper search of the base classes -- the
transitive closure of the search you've actually implemented.  I think
you would also need to stuff the newly created instancemethod into
__dict__ to avoid looking it up every time.

Jeremy




From HattiePastors at news.vex.net  Thu Nov  4 05:42:38 1999
From: HattiePastors at news.vex.net (Hattie Pastors)
Date: 4 Nov 1999 10:42:38 GMT
Subject: how to use pythondoc on Win32?
References: <7vrjqb$aju@mail.psy.uva.nl>
Message-ID: <7vrnuu$2koq$1@hub.org>

On Thu, 4 Nov 1999 10:32:18 +0100, Ionel Simionescu <ionel at psy.uva.nl> wrote:
>I did rename the extensionless 'pythondoc' to 'pydoc.py',

The secret for me was after renaming pythondoc to pydoc.py makes sure you
MOVE the pydoc.py OUT of the lib\pythondoc\ directory ... i found that
the program just won't run if you try to run the main script from
within the module directory. I don't know why, i never investigated. 
As soon as i finally got it working (after a few hours of frustration
like you have no doubt experienced) i was too happy to not to fiddle
with it anymore. It's been working ever since for me. 

It does say in the readme to move it into your 'path' somewhere. 
Of course the 'current directory' is in teh win32 path sort of, heh.
If my observation is right that the thing will not run from the 
module directory, the readme should be more specific on this account.

Great tool. Too bad the author hasn't had time to work on it much lately.





From a.eyre at optichrome.com  Thu Nov 25 07:50:02 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Thu, 25 Nov 1999 12:50:02 -0000
Subject: function as attr of subclass != method (???)
In-Reply-To: <81imde$3uk$1@nnrp1.deja.com>
Message-ID: <000401bf3743$980145e0$3acbd9c2@peridot.optichrome.com>

> >>> achild.method = lambda self:"nope"
> >>> achild.method()
> Traceback (innermost last):
>   File "<pyshell#102>", line 1, in ?
>     achild.method()
> TypeError: not enough arguments; expected 1, got 0

achild is an instance. You must assign the lambda as a member
of the child *class*, else it will be an unbound method.

e.g.

>>> class A:
...     pass
...
>>> b = lambda x: x
>>> b
<function <lambda> at 7f7bd0>
>>> a.b = b
>>> a.b
<function <lambda> at 7f7bd0>
>>> a.b()
Traceback (innermost last):
  File "<stdin>", line 1, in ?
TypeError: not enough arguments; expected 1, got 0
>>> a.b(1)
1
>>> A.b = b
>>> a=A()
>>> A.b
<unbound method A.<lambda>>
>>> a.b
<method A.<lambda> of A instance at 7f8d70>
>>> a.b()
<__main__.A instance at 7f8d70>

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From greg.ewing at compaq.com  Wed Nov 24 05:36:41 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Wed, 24 Nov 1999 23:36:41 +1300
Subject: automatic "lotjes trekken"
References: <81cqnv$2ip$1@vvs.superst.iae.nl> <81e3s2$j14$2@newshost.accu.uu.nl>
Message-ID: <383BBFB9.C9607961@compaq.com>

Martijn Faassen wrote:
> 
> Not at all. everybody here speaks Dutch anyway. (it's a conspiracy!)

Not me!

Uunlees-iit-meens-yoo-juust-haav-too-wrijt-aal-yoor-voowls-twijce,
Greg



From mal at lemburg.com  Fri Nov 26 12:21:41 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Fri, 26 Nov 1999 18:21:41 +0100
Subject: Newbie: raw_input and other input methods
References: <19991125.1481400@adam-aurag.penguinpowered.com> <81kfjc$1vg$1@vvs.superst.iae.nl>
Message-ID: <383EC1A5.99A4FAD5@lemburg.com>

Carel Fellinger wrote:
> 
> Hassan Aurag <aurag at geocities.com> wrote:
> 
> > Anyway, I used raw_input to be able to use readline and its nifty back
> > arrow, up arrow...
> 
> > However when I wanted to have a nifty prompt like the following one:
> 
> > raw_input("\033[0;34mInput: \033[0m"), I was surprised it got all
> > screwed up.
> 
> bash copes with this, so I dived into the source. And low and behold
> there it is: bash uses the readline special codes RL_PROMPT_START_IGNORE
> (='\001') and RL_PROMPT_END_IGNORE (='\002') to tell readline to 'ignore'
> the chars in between.  Using this info you can get what you want:)
> 
> e.g.
> >>>raw_input('\001\033[0:34m\002Input: \001\033[0m\002')
> 
> works like a charm, atleast here on my linux box

It even does the Right Thing for the Python prompt. Try this:

>>> sys.ps1 = '\001\033[0:34m\002>>> \001\033[0m\002'
>>> sys.ps2 = '\001\033[0:31m\002... \001\033[0m\002'
>>> if 1:
...     print 'looks cool :-)'
... 
looks cool :-)
>>>

The above adds a little color to your average Python prompt.

-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                    35 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/





From tendac at wanadoo.fr  Sat Nov  6 09:22:13 1999
From: tendac at wanadoo.fr (claude)
Date: Sat, 06 Nov 1999 14:22:13 +0000
Subject: eff-bot's The Standard Python Library
References: <19991105125124.A302546@vislab.epa.gov>
Message-ID: <38243995.94F2E17B@wanadoo.fr>

Ematter short story :
(sorry for my english, i am french)

1) i have tried to install it on a W98 personal computer,
and i saw the following message :

eMatter(c) Error
An error occured while preparing to display eMatter(c),
Please try again. If continue to have problems
contact Fatbrain.com Customer Service

=== I send an email to fatbrain customer service :
no answer
===

2) i have tried to install it on a Linux personal computer where
Nt4 SP5 is a "guest" reachable by VMWARE software (www.vmware.com).
All worked fine, i could read the book (acrobatreader V3) in spite of black
boxes appearing at the end of some paragraphes (acrobatreader errors 'Cs8' ???).
But when i left Nt4 session and started a new one with vmware and proceeded with
eMatter_The_Standard_Python_Librar.exe , i saw the following message :

There is a problem processing your saved registration information.
Please contact Fatbrain.com Customer Service

=== I send an email to fatbrain customer service :
 no answer
===

3)I have sucessfully installed eMatter_The_Standard_Python_Librar.exe on a true
Nt4 SP5
(not a linux-vmware one). All worked fine, i could read the book (acrobatreader
V3)
in spite of black boxes appearing at the end of some paragraphes (acrobatreader
errors 'Cs8' ???).
Then following the advice of the author of the book, i have upgraded acrobatreader
V3 to
acrobatreader V4, but now i CANT read the book :

eMatter(c) Error
An error occured while preparing to display eMatter(c), Please try again.
If continue to have problems contact Fatbrain.com Customer Service.

=== I send an email to fatbrain customer service :

You can go ahead and upload a new document on our site.  And then you can
delete the existing file that you are having problems with.  Please contact
me again if you have any further questions.

Thank you for using fatbrain.com
===

But in the meantime i cleaned up the "register base" (regedit)."Big Mistake".
I have had the following message

There are no new registration remaining for this license key, to view ematter
you will need to obtain an additional license key by purchasing this title from
fatbrain.com. If you believe you have reached this message in error , please
contatc Fatbrain.com customers service.

=== I send 2 emails to fatbrain customer service :
no answer.
===

Of course i have made a "big mistake", but i could not figure out
that this book was definitely tied up with Nt4.
If your NT box or w95/w98 box crash, the book is lost even if you
have made a backup of eMatter_The_Standard_Python_Librar.exe.,
this is a piece of nonsense, when you purchase a software it can
be install many times (Microsoft suite ...)

Best regards.




From wheinemanNOwhSPAM at uconect.net.invalid  Mon Nov 15 15:59:00 1999
From: wheinemanNOwhSPAM at uconect.net.invalid (wheineman)
Date: Mon, 15 Nov 1999 12:59:00 -0800
Subject: Need help dealing with spaces in NT paths.
References: <80pr76$jq5$1@nnrp1.deja.com>
Message-ID: <11f733ec.96958819@usw-ex0101-007.remarq.com>

Put quotes around the complete path. For example:

myExec = "C:\\Program Files\\Trash Can\\compactor.exe"



* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!




From sjturner at ix.netcom.com  Mon Nov  1 12:36:31 1999
From: sjturner at ix.netcom.com (Stephen J. Turner)
Date: Mon, 01 Nov 1999 12:36:31 -0500
Subject: Listbox.. ACTIVE-item is the one selected previously..
References: <ksemebjifu.fsf@telelogic.se>
Message-ID: <381DCF9F.F6D21C02@ix.netcom.com>

Andrew Markebo wrote:
> I want to have a listbox, and when the user clicks on one item in it,
> he gets the value selected 'published' somewhere.. with the small
> program included below.. If it is launched, and the user selects the
> letters in order I would assume the result to be "abcde", but what I
> get is: "aabcd", after the first selection, ACTIVE is lagged..
> 
> --------------------
> 
> z=None
> def click(event=None):
>    global z
>    print z.get(ACTIVE)
>    
> def test():
>    global z
>    frame=Frame(None)
>    frame.pack()
>    z=Listbox(frame)
>    z.pack()
>    z.bind("<Button-1>", click)
>    for i in ("a", "b", "c", "d", "e"):
>       z.insert(END, i)
> 
>    frame.mainloop()
> 
> test()

The code to activate the clicked element is found in the default listbox
class bindings (see listbox.tcl in your Tk distribution).  This requires
two changes to your script:

1.  The <Button-1> event specification should be changed to
    <ButtonRelease-1>, since activation occurs on button release.
2.  Your binding must be called after the class binding.

The second can be accomplished in two ways.  The first is to modify the
bindtags list for your listbox widget so that _all_ widget bindings
occur after class bindings:

   z.bind("<ButtonRelease-1>", click)
   tags=list(z.bindtags())
   mytag=tags[0]
   del tags[0]
   tags.insert(1, mytag)
   z.bindtags(tuple(tags))

This could cause problems, though, if you have other widget bindings
that you want to precede listbox class bindings.  In this case, what you
must do is associate your function with a new bindtag (e.g.,
"postclass"), and then insert the new bindtag in the listbox widget's
bindtags list after the class binding:

   z.bind_class("postclass", "<ButtonRelease-1>", click)
   tags=list(z.bindtags())
   tags.insert(2, "postclass")
   z.bindtags(tuple(tags))

Regards,
Stephen

--
Stephen J. Turner <sjturner at ix.netcom.com>



From emile at fenx.com  Mon Nov 22 07:10:47 1999
From: emile at fenx.com (Emile van Sebille)
Date: Mon, 22 Nov 1999 04:10:47 -0800
Subject: Order of keyword arguments
References: <38379233.ADE65CDB@corrada.com>
Message-ID: <008301bf34e2$9d3961e0$01ffffc0@worldnet.att.net>

Andres,

Perhaps you can adapt one of these ideas.

HTH


Emile van Sebille
emile at fenx.com
-------------------

#------start of code fragment

import traceback

def ordered1(opts ,**kws):
 print opts
 kws = eval('{'+opts+'}')
 print kws

def ordered2(opts ,**kws):
 print opts
 def convert_to_dict(**kws):
  return kws
 kws = eval('convert_to_dict(' + opts+ ')')
 print kws

def ordered3(**kws):
 tbs = traceback.extract_stack()

 for line in tbs:
  if line[3][:8] == 'ordered3':
   call_order = line[3][9:]

 print call_order
 print kws


ordered1 (opts="'c':1, 'b':2, 'a':1")
ordered2 (opts="c=1, b=2, a=1")
ordered3 (c=1, b=2, a=1)

#-------end of code fragment

----- Original Message -----
From: Andres Corrada <andres at corrada.com>
To: <python-list at python.org>
Sent: Saturday, November 20, 1999 10:33 PM
Subject: Order of keyword arguments


> Hi,
> Can one access the order in which keyword arguments were passed to a
> function? I want to call a function with the keyword syntax: f(
> foo='theFooString', bar='theBarString' ) and use the fact that foo was
> listed first so I can do something like:
>
> def f( **kw ):
>
> Process keyword arguments remembering the order in which the were used
> to call the function.
> Print first keyword
> Print second keyword
> .
> .
> .
>
> ------------------------------------------------------
> Andres Corrada-Emmanuel   Email: andres at corrada.com
> Owner                     http://www.corrada.com/mamey
> Mamey                     Phone: (413) 587-9595
> ------------------------------------------------------
>
> --
> http://www.python.org/mailman/listinfo/python-list
>
>





From billtut at microsoft.com  Wed Nov 24 16:26:13 1999
From: billtut at microsoft.com (Bill Tutt)
Date: Wed, 24 Nov 1999 13:26:13 -0800
Subject: Python and COM: a bug?
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8100DB915B2@RED-MSG-50>


> From: Gaetan Corneau [mailto:Gaetan_Corneau at baan.com]
> 
> 
> > I also don't think we try to optimize any calls. COM doesn't 
> > give one enough
> > information to even bother optimizing anything other than 
> > caching Name -> DispID mappings.
> 
> That's what I meant by "optimization".
> 

This optimization is allowed by COM. See the IDispatch::GetIdsOfName() docs
at: http://msdn.microsoft.com/library/psdk/automat/chap5_32cz.htm

"The member and parameter DISPIDs must remain constant for the lifetime of
the object. This allows a client to obtain the DISPIDs once, and cache them
for later use."

On the other hand, there is this as well:
"When GetIDsOfNames is called with more than one name, the first name
(rgszNames[0]) corresponds to the member name, and subsequent names
correspond to the names of the member's parameters. The same name may map to
different DISPIDs, depending on context. For example, a name may have a
DISPID when it is used as a member name with a particular interface, a
different ID as a member of a different interface, and different mapping for
each time it appears as a parameter."

If your COM object isn't following either of these rules then its broken.

The other intreresting thing to find out would be if the object supports
IDispatchEx.

Bill



From fdrake at acm.org  Thu Nov  4 16:59:43 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Thu, 4 Nov 1999 16:59:43 -0500 (EST)
Subject: Python Conference Hotel Change
In-Reply-To: <3824f510.12238077@news.demon.co.uk>
References: <199911032224.RAA10668@eric.cnri.reston.va.us>
	<3824f510.12238077@news.demon.co.uk>
Message-ID: <14370.463.892526.967619@weyr.cnri.reston.va.us>

Andy Robinson writes:
 > My memory may be fuzzy, but isn't it an even shorter stroll from there
 > to Bardo, one of the world's biggest microbreweries?

  A big microbrew?  Strange... hope it's good!  ;-)


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From gerrit.holl at pobox.com  Fri Nov 12 09:30:17 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Fri, 12 Nov 1999 15:30:17 +0100
Subject: Intelligent strptime?
In-Reply-To: <382AB016.D45727D8@strs.co.uk>
References: <19991110171256.A5196@optiplex.palga.uucp> <382AB016.D45727D8@strs.co.uk>
Message-ID: <19991112153017.B3867@optiplex.palga.uucp>

Ian Clarke explained why intelligent strptime is wrong:
> While I don't know what the application is, if you are processing user
> input I strongly recommend against doing this.  In a project I am
> working on the issue of whether to provide a "best guess" as to a date
> came up, but we decided that it was much better if a date had been
> entered incorrectly to furnish the user with an error, than attempt to
> correct that error automatically.  Sure in the short term it makes life
> a little easier for the user, but it also allows them to get used to the
> computer correcting their errors, and when the time comes that the
> computer cannot correct their error, or makes an incorrect guess, they
> will feel that there is something wrong ("I have always entered dates in
> MMDDYY format and it worked fine, why shouldn't it work now?").

Ah, ok. I'll force the user to type MM/DD and HH:MM...

regards,
Gerrit.
-- 
linuxgames.nl.linux.org	 All about games on Linux in Dutch (under construction).
www.nl.linux.org	       The Dutch resource for Dutch Linux documentation.
www.nl.linux.org/discoverb		    Learn foreign words and definitions.
www.nl.linux.org/~gerrit/asperger	 I'm an Asperger, go here for more info.
www.nl.linux.org/~gerrit			    About me and pictures of me.



From gmcm at hypernet.com  Wed Nov 10 12:11:46 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 10 Nov 1999 12:11:46 -0500
Subject: PyRun_SimpleString problem.
In-Reply-To: <80c5g9$1nm8$1@rtpnews.raleigh.ibm.com>
Message-ID: <1269868595-513164@hypernet.com>

"Arinte" writes:

> I am trying to run a python file from a dll.  When I run the file
> w/ PyRun_AnyFile it works fine.  But, when I try to run the code
> line by line w/ PyRun_SimpleString I get this error on function
> definitions.
> 
> def bottle(n):
>   File "<string>", line 1
>     def bottle(n):
>                 ^
> SyntaxError: unexpected EOF while parsing
>     if n == 0: return "no more bottles of beer"
>   File "<string>", line 1
>     if n == 0: return "no more bottles of beer"
> 
> Can I run files line by line if they define functions?  If so,
> how?

Nope. Can't do it with exec either. Even the interactive 
interpreter can't get away with it. You can only exec / eval a 
complete syntactic unit. You could look at a shell (eg, 
code.py) to see how's it done interactively, (normally append 
the line to a buffer, try running the buffer; if it worked, clean out 
the buffer). But if PyRun_AnyFile works, I wonder why you're 
even trying PyRun_SimpleString? Python would much prefer 
to work at the file level.

- Gordon



From mstenber at cc.Helsinki.FI  Mon Nov 29 01:30:39 1999
From: mstenber at cc.Helsinki.FI (Markus Stenberg)
Date: 29 Nov 1999 08:30:39 +0200
Subject: Anyone looked at pike?
References: <slrn84193k.6mc.kaleissin@rhiann.nvg.ntnu.no>
Message-ID: <al8aenx7qsw.fsf@sirppi.helsinki.fi>

kaleissin at nvg.org (Kaleissin) writes:
> Specifically: why they claim it's faster?

Shrug.. see below.

> http://pike.idonex.se/manual/tutorial/tut_introduction.html#introduction.3
> 
> Though, why anyone would want Yet Another C-Like Syntax is beyond me...<wink>

* It has static typing (which is much faster than dynamic used in Python,
* AFAIK)

Also, I'm not sure about how it's binding works and so on - but it seems to
be mostly "glue on top of C++" approach, not really dynamically
typed/late-bound language like Python. I personally would rather use C++(*)
really. Of course, I haven't been exposed for long. :)

I really wonder about similarities they mentioned on that page. It seemed
to me that it looked a lot more Perl-ish than Python-esque. Hmm. Maybe it
was all those braces and C(++) jargon they used.

> tal.

-Markus

P.S.
	from one example..

void find_song(string title)
{
	string name, song;
	.. (snip)
	if (string song=records[name][title])
            ^^^^^^ looks downright perverted to me :P

(*)
	I rather commit ritual suicide than code C++ in prolonged fashion.
	And I even (sort of) like C :P

-- 
The IBM Principle:		  
	Machines should work.  People should think.
The Truth About the IBM Principle:
	Machines don't often work, people don't often think.



From bowman at montana.com  Fri Nov 26 09:29:00 1999
From: bowman at montana.com (bowman)
Date: Fri, 26 Nov 1999 14:29:00 GMT
Subject: Python extensions and lcc-win32
References: <383E6378.4B10FDD6@europe.com>
Message-ID: <MOw%3.1498$XJ3.8798@newsfeed.slurp.net>

Kaloian Doganov <kaloian at europe.com> wrote:

>In fact, I'm looking for a non-MSVC solution to write Python extensions.
>So, any GCC or lcc-win32 tips are appreciated.

I've never tried lcc for Python extensions, but you might ask on the lcc ng.
There are several gcc or egcs ports to DOS/Windows.

www.delorie.com  is the djgpp compiler and tool set. It is DOS oriented and
a very robust set.

www.cygnus.com is the home of Cygwin. Cygwin is the tool of choice for ports 
from Linux to Dos/Windows. It pretty much replicates a mini-*nix environment 
under Windows.

There are links at that site to Mingw32, which is my favorite for Windows apps.
Ming is closely related to Cygwin. It was always the egcs port vs. gcc, but that
distinction is getting moot.

Any of these should work fine.






From Gaetan_Corneau at baan.com  Tue Nov 30 14:22:45 1999
From: Gaetan_Corneau at baan.com (Gaetan Corneau)
Date: Tue, 30 Nov 1999 14:22:45 -0500
Subject: Removing null items
Message-ID: <816010E2456BD111A48700805FBBE2EEFDF878@ex-quebec-u1.baan.com>

Take a look at the filter built-in function.

______________________________________________________
   Gaetan Corneau
   Software Developer (Quality Assurance Team)
   BaaN  Supply Chain Solutions  
   http://www.baan.com
   E-mail: Gaetan_Corneau at baan.com                  
   Tel: (418) 266-8252          
______________________________________________________
"Profanity is the one language all programmers know best"


> -----Original Message-----
> From: sragsdale at my-deja.com [mailto:sragsdale at my-deja.com]
> Sent: Tuesday, November 30, 1999 2:06 PM
> To: python-list at python.org
> Subject: Removing null items
> 
> 
> So I've got a very simple desire: to remove all false (I.E. 
> None, 0, '',
> []) items from a list.
> 
> As it is I've got this function which works fine, but it offends my
> sight.  Is there any better (or at least different) way to do this?
> 
> #################################################################
> # return non-null items from the first list
> def foo(lst):
>     rlist = []
>     for i in lst:
>         if i:
>             rlist.append(i)
>     return rlist
> 
> 
> Sent via Deja.com http://www.deja.com/
> Before you buy.
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 



From emuller at painewebber.com  Sat Nov 13 14:11:12 1999
From: emuller at painewebber.com (emuller at painewebber.com)
Date: Sat, 13 Nov 1999 19:11:12 GMT
Subject: Windows File Version Comparison....
Message-ID: <80kd4d$u14$1@nnrp1.deja.com>

I wrote a def that gets the file version of a windows exe/dll/cpl type
of file. It relies on filever.exe (part of the NT reskit) to be
available, but it should work for anyone looking to do this until the
win32ver module is finished. If you are looking for this module, it is
was under a '.dll and .exe version ...' thread. Anyway...

Once I did that, I could get the version of the file back, the next
thing I had to do was compare two versions and see which one was newer.
Since version strings can be '4.0.1.123'. I couldn't just convert them
to a float, and python thought that '4.0.1.6' was greater
than '4.0.1.42' for soem reason. So I wrote the following def to check
two version strings and return true/false....here it is...If anyone has
any problems with it, let me know, if anyone uses it let me know, if
anyone fixes a bug in it, let me know....

def compVer(ver1,ver2):
    #Checks to see if ver1 >= ver2
    import string
    vl1 = string.split(ver1,'.')
    vl2 = string.split(ver2,'.')
    while 1:
        if int(vl1[0]) > int(vl2[0]):
            return 1
        elif int(vl1[0]) == int(vl2[0]):
            del vl1[0]
            del vl2[0]
            if len(vl1) >= 1 and len(vl2) == 0:
                true = 1
                for each in vl1:
                    if int(each) <> 0:
                        true = 0
                return true
            elif len(vl1) == 0 and len(vl2) >= 1:
                true = 1
                for each in vl2:
                    if int(each) <> 0:
                        true = 0
                return true
            elif len(vl1) == 0 and len(vl2) == 0:
                return 1
            else:
                continue
        elif int(vl1[0]) < int(vl2[0]):
            return 0


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



From edwardm_painewebber at my-deja.com  Sat Nov 13 14:18:44 1999
From: edwardm_painewebber at my-deja.com (edwardm_painewebber at my-deja.com)
Date: Sat, 13 Nov 1999 19:18:44 GMT
Subject: Running a program as another user on Win32
References: <809ea4$ln$1@nnrp1.deja.com> <e11W3.33036$pf4.31242@news-server.bigpond.net.au>
Message-ID: <80kdii$uc8$1@nnrp1.deja.com>

Thanks,
I did not try the code (although I will), I did discover that before
you posted though. I used ntrights.exe (part of the NT Reskit) to grant
myself rights. The problem was I had to log off and on again for the
rights to take affect. This won't work for the application that I was
trying to use the program for. It looks like the security info for a
user is set during logon and not gotten again, at least not until the
next logon. I will try the code attached below once I have some time...

Great product so far Mark! I hope to be able to convince enough people
here to start using python vs. WinBatch. Especially since we can also
run python on our Solaris and AIX boxes (sans Win32 extension of
course)...

You know it amazes me....Most of the solaris people use korn shell
scripts, some use perl, the nt people use batch files and winbatch, the
AIX people (well god knows what they use)...If we all used python here
we could build a huge library of utilities and make large parts of it
work cross platform....It would be great......Ahhh..To dream....

:-)


In article <e11W3.33036$pf4.31242 at news-server.bigpond.net.au>,
  "Mark Hammond" <mhammond at skippinet.com.au> wrote:
> emuller at painewebber.com wrote in message <809ea4
$ln$1 at nnrp1.deja.com>...
> >I am writing a bunch of scripts that will need to run as
Administrator
> >(Domain Administrator) on Windows NT 4.0 workstation clients that
> >belong to an NT domain. I've tried using LogonUser, but I always get
> >the following: pywintypes.api_error: (1314, 'LogonUser', ' A required
> >privilege is not held by the client.'). I've also tried using
LogonUser
> >with the local Administrator account on my NT machine with no luck.
>
> You need to explicitely enable the privilege.  The following function
should
> do the job.  The documentation on CreateProcessAsUser is vague, and I
dont
> have time to test it.  It appears you need some or all of the
following
> privileges (all from the ntsecuritycon module):
SE_ASSIGNPRIMARYTOKEN_NAME,
> SE_INCREASE_QUOTA_NAME TOKEN_DUPLICATE, TOKEN_IMPERSONATE
>
> def AdjustPrivilege(priv, enable = 1):
>     # Get the process token.
>     flags = TOKEN_ADJUST_PRIVILEGES | TOKEN_QUERY
>     htoken = win32security.OpenProcessToken(win32api.GetCurrentProcess
(),
> flags)
>     # Get the ID for the privilege.
>     id = win32security.LookupPrivilegeValue(None, priv)
>     # Now obtain the privilege for this process.
>     # Create a list of the privileges to be added.
>     if enable:
>         newPrivileges = [(id, SE_PRIVILEGE_ENABLED)]
>     else:
>         newPrivileges = [(id, 0)]
>     # and make the adjustment.
>     win32security.AdjustTokenPrivileges(htoken, 0, newPrivileges)
>
> Mark.
>
>


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



From fredrik at pythonware.com  Fri Nov 12 10:03:51 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 12 Nov 1999 16:03:51 +0100
Subject: Tkinter and after
References: <382C2854.E9494EFE@lri.fr>
Message-ID: <009201bf2d1f$226b59c0$f29b12c2@secret.pythonware.com>

Stephane Conversy <conversy at lri.fr> wrote:
> It is said in Tkinter doc that after() is like a call to time.sleep()
> wich blocks Tkinter.

(what Tkinter docs?)

http://www.pythonware.com/library/tkinter/introduction/basic-widget-methods.htm
=> "alarm handlers and other non-event callbacks"

tells you how to register alarm handlers.

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
(news: just posted an errata for the 'first printing') -->




From shouldbe at message.com  Tue Nov 16 11:35:53 1999
From: shouldbe at message.com (Arinte)
Date: Tue, 16 Nov 1999 11:35:53 -0500
Subject: function override???
Message-ID: <80s0sd$1g4k$1@rtpnews.raleigh.ibm.com>

I am defining a python callback for my embedded app, but I want the script
writer to override my function.  So, basically what I want is a function
prototype.  Ex. here is what I am thinking.

;---fprot.py
def overrideme(self, str, someint)

;--fimp.py
import fprot   #I know this ain't right.
                          # I don't know how to import from other python
scripts yet.  Enlighten me?

def overrideme(self, str, someint)
    do_some_junk()







From kanzow at gmd.de  Fri Nov 26 07:30:50 1999
From: kanzow at gmd.de (Peter K.)
Date: Fri, 26 Nov 1999 13:30:50 +0100
Subject: Running apps with Zope 2?
References: <80v5g8$5a6$1@mulga.cs.mu.OZ.AU> <3834101A.FF39F64C@infercor.no> <8E83A8056PaCmAnRDLM@194.2.0.33> <slrn83jits.cj8.quinn@seniti.ugcs.caltech.edu>
Message-ID: <383E7D7A.1B48C7F1@gmd.de>

Hello,

I have written some web applications using the old Zope/Bobo version. I published
them using ZopeHTTPServer, like this:

File m.py:


class Application:
    "Application"
    def hello(self):
        "Hello"
        return "Hello World from my Zope Object"

app=Appl()


After starting ZopeHTTPServer (on hostname.domain.de) with
ZopeHTTPServer -p 1234 m.py
I can access the Application object from the web:
requesting  http://hostname.domain.de:1234/app/hello
gets "Hello World from my Zope Object".

This is what I want! But what must I do to achieve this in Zope 2? I installed it
recently and can't find it in the documentation. How can I publish my old
applications with the new Zope? Where do I have to put m.py? What do I have to do
in the management screen? Which URL will get me the "Hello World..." message? Do I
have to install m.py as a Product? How?

Thank you very much

Peter Kanzow, Germany





From gmcm at hypernet.com  Sat Nov 13 21:48:55 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sat, 13 Nov 1999 21:48:55 -0500
Subject: method overriding trick
In-Reply-To: <382DFF59.51F186A6@x.x>
Message-ID: <1269574764-2892811@hypernet.com>

Eric Jacobs wrote:

> Phil Mayes wrote:
> > 
> > I use a method that doesn't remove the clunkiness, but localises the
> > hierarchy to one line:
> > 
> > _Parent = basedlg.BaseDlg
> > class AddressDlg(_Parent):
> >     def __init__(self, param):
> >         self.local = param
> >         _Parent.__init__(self, 2345)
> > 
> > By using the same name "_Parent" in all modules, less thinking is needed,
> > and cut'n'paste coding (what, me?) works better too.
> 
> You wouldn't be able to reassign the _Parent variable to define another
> class elsewhere in the same module, though, because the functions defined
> this way will always look up the _Parent from the module namespace at
> execution time. 

Unless you've nested your class statement inside a def or 
another class, the class object will get built when the module 
is loaded. So what the _Parent variable holds once your code 
starts executing doesn't matter - it is only meaningful at the 
time the "class" statement is executed.

Sure you can reassign it, but it will only mean something if 
you do it at the top level.

OTOH, I find that it's rare for a base class and it's derived 
classes to have the same __init__ signature, so for me the 
trick is rarely useful.

- Gordon



From endgamer71 at hotmail.com  Tue Nov  2 06:07:28 1999
From: endgamer71 at hotmail.com (Endgamer)
Date: Tue, 02 Nov 1999 11:07:28 GMT
Subject: Newbie question : Info on Python under DOS/Windows
Message-ID: <381ec4c4.7451128@212.225.36.99>

Okay, I'm really new to Python, and have limited experience of real
programming (yes, the age-old story....crippled by BASIC at about the
age of 10 <g>). I'm learning from texts, this froup and O'Reilley's
"Programming Python" and as a starting project I'm writing a Y2K
mini-fix for Wintel machines - a program which prompts the user for
the current date and feeds it into "date", for machines which won't
hold the date on reboot after the millenium. 
My question is: I've found it hard to locate resources or  information
on Python under DOS/Windows - are there any special groups, sites or
FAQs/Manuals with general information? Much as I hate MS environments,
in my line of work I have to work with them almost constantly (back on
my lovely Linux machine whenever I get a chance <g>) and I suspect
I'll need this information again...


Endgamer
Spider Warrior

(Endgamer71 at spamtrap.hotmail.com)
#####################################
Not a quote, just an on-the-spot observation:
You know you're a geek when your PC costs exactly
15 times as much as your stereo. And gives better
sound quality.<sigh>
                              Endgamer
#####################################



From you at somehost.somedomain  Tue Nov 30 11:15:04 1999
From: you at somehost.somedomain (****)
Date: 30 Nov 1999 16:15:04 GMT
Subject: NEWBIE
References: <81m7m3$82$1@news.qub.ac.uk> <38428160.56B1D46D@gssec.bt.co.uk>
Message-ID: <820t68$asr$1@news.qub.ac.uk>

In article <38428160.56B1D46D at gssec.bt.co.uk>, alan.gauld at gssec.bt.co.uk 
says...
>
>Stuarty at excite.co.uk wrote:
>> I have just recently started using Python, after the first day I was 
further
>> on than I am after a semester of C programming(were at arrays...).
>> tutorial by Giudo van Rossum was very useful, but I attempted printing out 
the
>> reference library..there are now less considerably less trees in the world.
>
>So are you asking a question?
>....or just boasting? ;-)
>
>Alan g.
>-- 
>=================================================
>This post represents the views of the author 
>and does not necessarily accurately represent 
>the views of BT.

Pointing out how easy Python is to use, to a relative beginner :-)




From oli at andrich.net  Tue Nov 23 11:58:32 1999
From: oli at andrich.net (Oliver Andrich)
Date: Tue, 23 Nov 1999 17:58:32 +0100
Subject: Asyncchat Module Documentation and some asyncchat based modules
Message-ID: <19991123175832.A26756@gothic.andrich.net>

Hi,

after I spend most of the last weeks implementing some asyncchat based clients
for common used protocols (at least I use the quite common ;-) and a asyncchat
based POP3 server. I like to ask if somebody already started to document this
module for the library reference, otherwise I am willing and happy to do it.
I am so impressed by asyncore and asyncchat, that I like to do it. ;-)

And I am looking forward to do a first release of my package of asnyc* based
clients in a week or two. And my POP3 server (100% pure Python) in about three
weeks. I have to to do some code cleanups, security checks and performance
tuning, but at the moment the POP3 server runs fine in my testbed (simulating
the 16.000 users I have on my production machine) and it outperforms cucipop
without any problems.

So, who do I have to contact to supply documentation for asynchat?

Best regards,

    Oliver

-- 
Oliver Andrich, KEVAG Telekom GmbH, Schlossstrasse 42, D-56068 Koblenz
Telefon: 0261-3921027 / Fax: 0261-3921033 / Web: http://rhein-zeitung.de




From bjorn at roguewave.com  Mon Nov 29 16:44:31 1999
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Mon, 29 Nov 1999 14:44:31 -0700
Subject: Python bugs
Message-ID: <51C9F3C392D0D111A15600A0C99E64A9AF569E@localhost>

>     Bjorn> Which is a good thing, but shouldn't preclude it 
> from being used
>     Bjorn> as a dictionary in the update statement (although the fix
>     Bjorn> probably have to be to Dict.update rahter than os.environ).
> 
> You are correct.  It would have to make some assumptions 
> about how to get a
> dict from an instance object.  Should it access the 
> instance's "__dict__"
> attribute or (in this case) the "data" attribute or try to an 
> "as_dict"
> method?  Since no such well-defined interface currently exists, you're
> probably better off in the short term simply executing
> 
>     d.update(os.environ.data)

How about:

	def update(self, other):
		... regular stuff
		if
other_is_an_instance_or_something_else_we_dont_know_about:
			for k in other.keys():
				self[k] = other[k]

That should be well defined enough, or raise the appropriate exception if it
can't be done.(?)


>     Bjorn> Cool (do you know if you can do db.items() also? 
> -- found that
>     Bjorn> one five minutes ago...)
> 
> I don't see it, and I doubt it's going to get added any time 
> soon.  One
> motivation for using on-disk mappings such as provided by 
> bsddb and (g)dbm
> is that they can be really huge (hundreds of thousands or 
> millions of keys
> are not out of the question).  Asking for values() or items() in these
> situations is much more likely to get you into memory trouble than for
> simple dicts which tend to be smaller.

[workarounds snipped]

Yes, all of these are easy to work around if you really need to -- the point
being you shouldn't need to <0.1 wink>

It's true that on-disk mappings can get huge.  It's also true that sometimes
they're just used for convenient persistence of moderately sized data.  I
think it's generally a bad idea to try to prevent people from potentially
shooting themselves in the foot :-)

-bjorn



From arcane at znet.com  Sat Nov  6 16:34:10 1999
From: arcane at znet.com (Robert Cain)
Date: Sat, 06 Nov 1999 13:34:10 -0800
Subject: python: something completely different
References: <3820C0B8.62B2@iowegian.com>
Message-ID: <38249ED2.BB7EA1DE@znet.com>

Back to the subject at hand (Python in the DSP universe) has anyone
investigated the possibility of conversion, compilation or whatever of
MatLab to Python?  I'm naive WRT to both at this point because 1) MatLab
is too expensive to consider, 2) MatLab costs too much, and 3) I've just
been turned on to Python as a result of the original post in this
comp.dsp thread.  Looks like it could be the wavelet of the future.


Peace,

Bob

-- 

"Things should be described as simply as possible, but no simpler."

                                             A. Einstien



From pinard at iro.umontreal.ca  Mon Nov  1 19:39:16 1999
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 01 Nov 1999 19:39:16 -0500
Subject: Improved struct module
In-Reply-To: wtanksle@hawking.armored.net's message of "1 Nov 1999 16:43:58 GMT"
References: <000001bf15fc$3397ee00$ba2d153f@tim> <0E16861EE7BCD111BE9400805FE6841F0BEB2650@c1s5x001.cor.srvfarm.origin-it.com> <oqso2phk2j.fsf@titan.progiciels-bpi.ca> <slrn81s9mj.d51.wtanksle@hawking.armored.net>
Message-ID: <oqk8o1hgln.fsf@titan.progiciels-bpi.ca>

wtanksle at hawking.armored.net (William Tanksley) ?crit:

> >> I must admit that I am also one of those lazy bastards that uses his
> >> newsreader's "find any messages by Tim Peters"-function to locate
> >> interesting threads.

> >So do I, but even more lazily!  Gnus bolds Tim's messages automatically
> >for me, when preparing the threaded summary for the Python mailgroup. :-)

> Ah, but does your newsreader automatically delete all the other messages?
> Hah!  Some time-saver it is.

You forgot the smileys (or <winkles>)!  No, surely not.  There are a lot
of interesting contributions, opinions and attitudes.  I try to at least
throw a quick eye on everything.  But it's not easy: after I re-established
a mail gateway machine around here, which has been down for around five
days, I got a flurry of maybe 6000 messages, and my Python mailgroup only
got over 1000 for a while.  Now back below 100, but some messages are very
difficult to get rid of, as they are so informational that I hardly want
to even file them away for later study...  You are too interesting, guys!

> I used to use nn (NoNews).  Now THAT was a good newsreader -- or news
> ignorer as the case may be.  [...]

For users, this is cake, and always have been so far that I know.  On the
other hand, (and I do not know where `nn' stands nowadays), I do remember
that many years ago, for a system administrator, `nn' was sometimes turning
into pure frantic nightmare.  When things were starting to go wrong,
rebuilding indices and everything was taking _days_ of machine time
(machines were also much slower than in these days.)  Many years ago,
I've been hired by a group that gave me the mandate of finding a way to
gracefully untangle `nn' from their system, while keeping their users happy.
It was a difficult undertaking for the poor little me! :-)

> In most newsreaders the default is to read every message; in NN the
> default is to ignore every message.

I'm not going to give in a newsreader war, and I guess that best ideas just
flood into all newreaders after a while.  Gnus is immensely configurable,
and I'm confident that users could do almost whatever we want with it.
I use it to read my mail, search databases, and a few other things.  Even,
sometimes, not often, to read news :-).  On the user side, I would easily
bet that `nn' is faster, but on reasonable machines, Gnus is very workable.

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





From greg.ewing at compaq.com  Tue Nov  2 07:23:03 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Wed, 03 Nov 1999 01:23:03 +1300
Subject: Comparison between Python and "Ruby"
References: <1270663667-24391282@hypernet.com>
Message-ID: <381ED7A6.D7D7FB13@compaq.com>

Gordon McMillan wrote:
> 
> That's the Perler's war cry. Snort. As if there was only one
> way to have more than one way to do it.

Okay, so the Python war cry should be TMTOWTHMTOWTDI
(There's More Than One Way To Have More Than One Way
To Do It).

Greg



From a.eyre at optichrome.com  Mon Nov 29 08:42:35 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Mon, 29 Nov 1999 13:42:35 -0000
Subject: help with opening html files
In-Reply-To: <028601bf3a5d$31abdea0$f29b12c2@secret.pythonware.com>
Message-ID: <000a01bf3a6f$982ead80$3acbd9c2@peridot.optichrome.com>

>> Can anyone tell me how to open a html file from within python in
>> windows.

> html files are text files, and can be handled
> in the usual way.  see chapter 7 in the tutorial
> for more info:

Unless you want to open one from the web, in which case, have a look
at httplib.

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From mhammond at skippinet.com.au  Fri Nov 26 07:31:16 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Fri, 26 Nov 1999 12:31:16 GMT
Subject: dir() and propertly list NEWBIE
References: <199911261153.LAA07241@cobweb.quantisci.co.uk>
Message-ID: <o4v%3.6690$VA6.32052@news-server.bigpond.net.au>

Steve M S Tregidgo wrote in message
<199911261153.LAA07241 at cobweb.quantisci.co.uk>...
>Donn Cave wrote:
>> Quoth tiddlerdeja at my-deja.com:
>> [snip]
>> | Similarly, If I can get an objects functions, is it possible then
>get a
>> | list of the possible parameters for that function?
>>
>> No!  Wouldn't that be nice?
>Take a look at the function's attribute 'func_code'.  Within this

Or even better - use IDLE or Pythonwin - you can (usually :) type the
function name and have a short description and the parameter list pop up.

Mark.






From m.faassen at vet.uu.nl  Tue Nov 16 06:11:41 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 16 Nov 1999 11:11:41 GMT
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <JlFX3.3659$s87.67930@dfiatx1-snr1.gtei.net> <80oqn9$ocq$2@newshost.accu.uu.nl> <oLUX3.286$yA4.4821@dfiatx1-snr1.gtei.net> <80p9su$ppo$2@newshost.accu.uu.nl> <80piru$13rl$1@hub.org>
Message-ID: <80re5d$55a$1@newshost.accu.uu.nl>

T. wrote:
> On 15 Nov 1999 15:46:38 GMT, Martijn Faassen <m.faassen at vet.uu.nl> wrote:
>>mature enough, Parnassus can have a flag somewhere 'distutils compatible'
>>or something like that for each module/program in its database.

> Yeesh! Got about a hundred times more hits on my web site today than i've
> ever gotten. And now i'm posting ten times more usenet messages in one
> day than i usually do in a week or a month or more! Grumble grumble. (-;

You're hard to keep track of too, your listed email address keeps changing! 

Getting-curious-about-the-person-with-the-mysteriously-changing-name-ly yours,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?



From aahz at netcom.com  Wed Nov 10 12:15:41 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 10 Nov 1999 17:15:41 GMT
Subject: Python company worth $100M
References: <38295BC2.13E2E31F@digicool.com> <033201bf2b73$ad1e49e0$f29b12c2@secret.pythonware.com>
Message-ID: <80c97t$r6$1@nntp9.atl.mindspring.net>

In article <033201bf2b73$ad1e49e0$f29b12c2 at secret.pythonware.com>,
Fredrik Lundh <fredrik at pythonware.com> wrote:
>
>  [ Searchbutton.com ]
>and if you look at the technical credits page, what do you
>find?
>
>    http://www.searchbutton.com/misc/technicalCredits
>
>asynclib again, of course ;-)

Actually, we're moving away from asyncore in favor of threads.  Unless
we're missing something obvious, asyncore just doesn't do well with
compute-heavy processes.
--
> I wrote:
> > a slightly hairer design issue is what combinations
> > of pattern and string the new 're' will handle.
> 



From ivanlan at callware.com  Sat Nov  6 13:23:54 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Sat, 06 Nov 1999 11:23:54 -0700
Subject: making the Python case to my dept.
References: <9911051740130D.02530@william.jinsa.org> <3823FC43.CF7FBD71@python.net> <801p3q$q01$1@nntp2.atl.mindspring.net>
Message-ID: <3824723A.5DE35FE5@callware.com>

Hi All--

Aahz Maruch wrote:
> 
> First of all, I want to second Patrick's suggestion that this be FAQ'd.
> I have one nit, though:
> 
> In article <3823FC43.CF7FBD71 at python.net>,
> Thomas A. Bryan <tbryan at python.net> wrote:
> >
> >Still, few people are knocking down my door for Python programmers
> >while there are plenty of jobs for C++/Java/Delphi programmers.
> 
> I believe that there are more jobs for good Python programmers than
> people with the skills to fill them.  My company is certainly having
> difficulties (jobs in Mountain View! ask me!), and having HP Cupertino
> advertising for 10-12 Python positions only increases the difficulty.
> 
> I think we've come to the knee in the curve where Python has achieved
> critical mass and will get a lot of popular attention over the next
> couple of years.  Few of us need worry about job security.
> -

I really hope that Aahz is correct.  As most of you know, I'm stuck out
here in Utah, which is widely known to be significantly behind *any*
curve you want to name.

If there's a curve coming to Utah, I sure hope it will be one with jobs
for Pythonistas, because there certainly aren't *any* here now.

<knee?-we-haven't-even-seen-the-toe!>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From aa8vb at yahoo.com  Mon Nov  1 11:12:31 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Mon, 1 Nov 1999 11:12:31 -0500
Subject: Regex problem with multiline matches
In-Reply-To: <14361.62318.90382.177379@dolphin.mojam.com>
References: <19991029150848.A106981922@vislab.epa.gov> <14361.62318.90382.177379@dolphin.mojam.com> <0e2601bf2245$126cef00$0100a8c0@rochester.rr.com> <19991029150848.A106981922@vislab.epa.gov> <14361.62318.90382.177379@dolphin.mojam.com>
Message-ID: <19991101111231.B133876404@vislab.epa.gov>

Skip Montanaro:
 |Randall Hopper:
 |    >>>> print re.compile( "b(.)", re.M ).match( "aa\nbb" )
 |    None
 |
 |The match method anchors the search at the beginning of the string.  Try the
 |search method instead:
 |
 |    >>> print re.compile( "b(.)", re.M ).search( "aa\nbb" )

Thanks.  I should have caught that in the docs.  'match' just didn't
register as really 'match_prefix' or 'match_at_beginning_of_line'.

Darrell:
 |Or
 |
 |>>> print re.compile( "(?s).*?b(.)").match( "aa\nbb" )

Hey that's slick!  Learn something new every day.  Just change the
interpretation of period so it can eat newlines too.

Thanks guys!

-- 
Randall Hopper
aa8vb at yahoo.com



From bernhard at alpha1.csd.uwm.edu  Tue Nov 30 13:21:56 1999
From: bernhard at alpha1.csd.uwm.edu (Bernhard Reiter)
Date: 30 Nov 1999 18:21:56 GMT
Subject: XPython (X11)
References: <383E5350.1E07536A@xrce.xerox.com> <slrn83tj53.k7h.bernhard@alpha1.csd.uwm.edu> <81ms2n$126$1@essle.valdyas.org> <slrn840dt7.mov.bernhard@alpha1.csd.uwm.edu> <81vs99$g8$1@essle.valdyas.org>
Message-ID: <slrn8485e4.cn.bernhard@alpha1.csd.uwm.edu>

On 30 Nov 1999 06:53:29 GMT, Boudewijn Rempt <boud at rempt.xs4all.nl> wrote:
>Bernhard Reiter <bernhard at alpha1.csd.uwm.edu> wrote:
>> On 26 Nov 1999 20:54:47 GMT, Boudewijn Rempt <boud at rempt.xs4all.nl> wrote:
>>>Bernhard Reiter <bernhard at alpha1.csd.uwm.edu> wrote:

>> And I didn't mention pyGNOME, why should I mention to PyKDE?
>Because the two go together - you did mention PyGTK.

True. But I was thinking GUI libraries. The gnome libraries are
If I am still informed correclty a lot more then that.

>> To create an application tied to KDE or GNOME is a mistake in my
>> eyes anyway. <wink>

>Oh, yes, people will agree with you. 
No sadly enough, they don't. They write every application a couple of
times. At least once for GNOME and KDE and then there is always the forgotten
version, makes at least a trippe effort.
(I know I know there are reasons to reimplement sometimes, I am not arguing
these cases.)

>I've one regular correspondent
>who berates me continually for not writing my apps to run inside
>a browser. 

Well a browser is a very limited application development platform.
I think it is overrated and born because people (esp.) made it hard to
develop portable software.

>I don't really care, though - sometimes I need an
>application that's integrated with the rest of my work environment,
>and sometimes I can do with something standalone.

People are so trained in bad user interfaces that they don't 
value got ones. And the web is not a good user interface for a lot
things.

>> people-will-love-it-any-way-ly, yours
>They do, they do, too... ;-).

Perceiving-it-as-a-two-edge-sword-ly, yours
	Bernhard


-- 
Research Assistant, Geog Dept UM-Milwaukee, USA.  (www.uwm.edu/~bernhard)
Free Software Projects and Consulting 		         (intevation.net)  
Association for a Free Informational Infrastructure            (ffii.org)



From a.eyre at optichrome.com  Tue Nov 16 08:45:01 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Tue, 16 Nov 1999 13:45:01 -0000
Subject: Platform independent? (Was: Drawing (directed) graphs)
In-Reply-To: <14371.2783.885109.27130@weyr.cnri.reston.va.us>
Message-ID: <004101bf3038$c7d96c00$3acbd9c2@peridot.optichrome.com>

>> This is why it annoys me that some of the scripts in the
>> standard Python library are not platform independent, but
>> rely on non-portable features of posix.

> If you know of ways to reduce the platform-dependence of these,
> please feel free to submit patches!

I'm sure a lot of the base SocketServer stuff could be rewritted
to use the select module (which will port to Win32).

The other alternative (if this is not possible) is to move any
scripts in Python's Lib/ directory which are dependent on posix
into a subdirectory called plat-posix/ much like the plat-win/
stuff.

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From wtanksle at hawking.armored.net  Sat Nov 27 14:41:28 1999
From: wtanksle at hawking.armored.net (William Tanksley)
Date: 27 Nov 1999 19:41:28 GMT
Subject: Type/Class Unification in Ruby
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp> <7vjtrc$hmm$1@newshost.accu.uu.nl> <38204614.2EAC6051@compaq.com> <5ld7trtak5.fsf@eric.cnri.reston.va.us> <3826D63C.2731BB1F@compaq.com> <38328948.9E35556F@compaq.com> <m3d7t9v8lr.fsf@qiao.localnet> <38356C74.F9E4248D@compaq.com> <14397.64943.791678.643871@anthem.cnri.reston.va.us>
Message-ID: <slrn840d4d.bt2.wtanksle@hawking.armored.net>

On Thu, 25 Nov 1999 22:25:35 -0500 (EST), Barry A. Warsaw wrote:
>>>>>> "GE" == Greg Ewing <greg.ewing at compaq.com> writes:

>    GE> A middle ground might be to allow inheriting from any number
>    GE> of user-defined classes, but at most one built-in type.

>And in fact, this is exactly the JPython restriction.  Because Java
>has only single inheritance, a JPython class can only have a single
>Java base class, but any number of pure Python base classes.

Another possibility would be to borrow from Sather by making all
inheritance actually act like interface inheritance with delegation.

In other words, JPython creates an interface for every class, and when you
inherit you actually implement and then add a member data thingie.

>-Barry

-- 
-William "Billy" Tanksley



From wtanksle at hawking.armored.net  Sat Nov 27 14:45:46 1999
From: wtanksle at hawking.armored.net (William Tanksley)
Date: 27 Nov 1999 19:45:46 GMT
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <38397696@news.xtml.co.uk> <221119992150405215%zippy@cs.brandeis.edu> <383a8909@news.xtml.co.uk>
Message-ID: <slrn840dcf.bt2.wtanksle@hawking.armored.net>

On Tue, 23 Nov 1999 12:38:36 -0000, Nick Trout wrote:
>How do you write large/complicated applications in Lisp, the syntax with all
>those brackets drives me mad!

Just ignore them -- pretend you're writing Python.  Let the indentation
guide you.  If you pay attention to the parens you WILL go mad.

>I've never heard of this clever type checking. It sounds like a very good
>idea. It would be nice to be able to choose whether to have typing on or
>off. I know you can run a Lint checker over Python but whats a compiler
>for?!

Lisp is truly amazing.  Anything that's cool in any language existed in
Lisp before, and probably is still more usable in Lisp.

>N

-- 
-William "Billy" Tanksley



From helge.hess at mdlink.de  Tue Nov 30 13:29:04 1999
From: helge.hess at mdlink.de (Helge Hess)
Date: Tue, 30 Nov 1999 19:29:04 +0100
Subject: Python Make
References: <3843E558.8AC91A97@strs.co.uk>
Message-ID: <38441770.80C33C93@mdlink.de>

Hi,

Ian Clarke wrote:
> 
> Hi, just a quick idea I had that I wanted to share with the newsgroup.
> Has anyone considered using Python as a replacement for the "make"
> utility?  A module could be provided which did several small things like
> testing modification times of a list of source files against a
> destination file to see whether an update was required.

YES! I had the same idea :-)

> I am prompted to suggest this after spending yet another half-day trying
> to debug a Makefile.
> 
> What do people think?

I think it is a pretty cool idea. The negative aspect is that one needs
python for compilation then (eg you can't compile python using 'python'
makefiles).

I also wonder about a nice Python syntax for doing this. Did you think
about this issue already ? A simple makefile should look as easy as a
usual Makefile, maybe like this:

  a = Source('a.c')
  b = Source('b.c')
  c = Source('c.c')

  def compileC(source):
     gcc(source)
  
  all = { 'dependsOn': ( a, b, c),
          'action': compileC }

Hm, looks somewhat strange and a powerful 'make' framework needs to be
thought out first.

Greetings
  Helge
-- 
SKYRIX-OS  Web Application Environment - http://www.skyrix.com
SKYRIX-IAS Groupware Application Suite - http://www.skyrix.com



From thomas at xs4all.nl  Fri Nov  5 04:22:33 1999
From: thomas at xs4all.nl (Thomas Wouters)
Date: Fri, 5 Nov 1999 10:22:33 +0100
Subject: __str__ vs. __repr__
In-Reply-To: <000001bf275c$02370d20$1d2d153f@tim>; from tim_one@email.msn.com on Fri, Nov 05, 1999 at 02:04:30AM -0500
References: <199911041325.IAA14001@eric.cnri.reston.va.us> <000001bf275c$02370d20$1d2d153f@tim>
Message-ID: <19991105102233.I29354@xs4all.nl>

On Fri, Nov 05, 1999 at 02:04:30AM -0500, Tim Peters wrote:

> and Python's European groupies didn't bitch enough about
> repr(euro_string) being an unreadable backslashed mess.

We're too busy bitching about the idiocy of people who expect to change
a country's (usually) ancient currency, habits and not to mention every
keyboard, font and accountancy software in the world in a matter of years.
Or perhaps we just dont _care_ about the euro symbol. Your Euro symbol, by
the way, Tim, ended up as '?' in my font -- but possibly because isprint()
returned 0 and my mail-reader therefor refused to print it ;)

Oh, and we bitch about why they had to rotate the Quake II logo 90 degrees,
too. I think acceptance of the logo would have been easier if it'd been
sitting up straight.

ISO-8859-15-aka-Latin-9'ly y'rs,
-- 
Thomas Wouters <thomas at xs4all.net>

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



From levicook at earthlink.net  Tue Nov 30 12:20:06 1999
From: levicook at earthlink.net (Levi Cook)
Date: Tue, 30 Nov 1999 11:20:06 -0600
Subject: ? on working with http redirects in htmllib
Message-ID: <821146$266$1@ash.prod.itd.earthlink.net>

Does anyone know how to process and follow a redirect in python?

I can easily tell if I've recieved the 301 or 302 errcode, but I cannot for
the life of me figure out how to get the URL of the new destination.

Any help is greatly appreciated.
- Levi Cook

##### code snippet that i use to retrieve pages...

import httplib, urllib

address = 'http://www.python.org/doc/current/lib/node227.html'

URL2Fetch = urlparse.urlparse( address )

# fetch page
h = httplib.HTTP(URL2Fetch[1])
h.putrequest('GET', URL2Fetch[2])
h.putheader('Accept', 'text/html')
h.putheader('Accept', 'text/plain')
h.endheaders()
errcode, errmsg, headers = h.getreply()

# script only knows how to process html
if not headers.gettype() == 'text/html':
     raise redirect(address)

f = h.getfile()
data = f.read() # Get the raw HTML
f.close()







From jae at ilk.de  Sun Nov  7 16:28:31 1999
From: jae at ilk.de (Juergen A. Erhard)
Date: Sun, 07 Nov 1999 22:28:31 +0100
Subject: pythonistas or pythoneers?
Message-ID: <07111999.2@sanctum.jae>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

I'm confused: are we pythoneers or pythonistas?

I tend to go for pythoneers... sounds more pythonic.

Or are these two levels of experience?  If so, what's higher?

Confused,

Bye, J

PS: Of courthe, my thongue wath firmly lodthed in my theek while I
wrote thith...

- -- 
J?rgen A. Erhard      eMail: jae at ilk.de      phone: (GERMANY) 0721 27326
    MARS: http://members.tripod.com/~Juergen_Erhard/mars_index.html
           "Outside of a dog, a man's best friend is a book;
           inside of a dog, it's very dark." --  Groucho Marx
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.0 (GNU/Linux)
Comment: For info see http://www.gnupg.org

iEYEARECAAYFAjgl7v0ACgkQN0B+CS56qs1dNACeOdnrX6Iu5upos7+g6NkbPaWl
2aEAnie43H0TH9tAJRB1tc6nfl0LnfQO
=QUO3
-----END PGP SIGNATURE-----



From fdrake at acm.org  Wed Nov 24 09:22:02 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Wed, 24 Nov 1999 09:22:02 -0500 (EST)
Subject: whrandom seed question
In-Reply-To: <000601bf3645$a8dc0ca0$f82d153f@tim>
References: <81eu22$411al$1@umbc7.umbc.edu>
	<000601bf3645$a8dc0ca0$f82d153f@tim>
Message-ID: <14395.62602.126765.991856@weyr.cnri.reston.va.us>

Tim Peters writes:
 > Don't use module whrandom.  Use module random.  The former is an
 > implementation detail of the latter, and follows the fine scientific
 > tradition of using incomprehensible calling conventions <0.5 wink>; whrandom
 > isn't meant to be used directly; random is; someday the docs will catch up

  For the record, I intend to have the docs for random and whrandom
fixed up before the next documentation release.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From greg.ewing at compaq.com  Mon Nov  8 06:14:12 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Tue, 09 Nov 1999 00:14:12 +1300
Subject: A newbie dictionary question
References: <byr9i42us9.fsf@bolzano.math.ethz.ch>
Message-ID: <3826B084.20F9B58D@compaq.com>

Jan Vroonhof wrote:
> 
>     def __cmp__(self,other):
>         return self.nospaces() == other.nospaces()

A __cmp__ function is supposed to return -1, 0 or 1 corresponsing
to less than, equal to or greater than. Try it this way:

     def __cmp__(self,other):
         return cmp(self.nospaces(), other.nospaces())

Greg



From robin at alldunn.com  Fri Nov 19 12:43:28 1999
From: robin at alldunn.com (Robin Dunn)
Date: Fri, 19 Nov 1999 09:43:28 -0800
Subject: Building wxPython [was Re: wxPython and DISLIN work nicely together]
References: <80s0b4$g5n@mail.psy.uva.nl> <19991117090214.B807152@vislab.epa.gov> <017a01bf31f8$6bbea590$1a25d2d1@jenkondev.com> <19991119064944.A996886@vislab.epa.gov>
Message-ID: <002301bf32b5$97a5d440$0301a8c0@easystreet.com>

What is the output of "wx-config --cflags" ?  It should have the right
include directories for getting all the wxWindows headers.

Oh wait a minute, it says "/bin/sh: wx-config:  not found" so it looks like
you didn't do a "make install" for wxGTK.

Robin

----- Original Message -----
From: Randall Hopper <aa8vb at yahoo.com>
To: Robin Dunn <robin at alldunn.com>
Sent: Friday, November 19, 1999 3:49 AM
Subject: Re: wxPython and DISLIN work nicely together


> Robin Dunn:
>  |>
>  |> If only wxPython would build on IRIX...
>  |>
>  |
>  |Did you have build issues with wxWindows or did you get as far as trying
to
>  |build wxPython?  If so, what were the troubles?
>
> I've attached a message that describes the problem I had with it when I
put
> it aside.  Tips appreciated.
>
> Randall
>




From g_will at cyberus.ca  Sat Nov 27 08:59:43 1999
From: g_will at cyberus.ca (Gordon Williams)
Date: Sat, 27 Nov 1999 13:59:43 GMT
Subject: Pythonwin Reset
Message-ID: <01bf38df$8127e0c0$3a42c3d1@Cyberus.cyberus.ca>

Hello All,

Is there a way to reset Pythonwin IDE between runs.  I want to eliminate
any rogue objects or modules that are still in memory from the previous run
of the program.

Thanks

Gordon Williams



From jam at quark.emich.edu  Tue Nov 16 07:23:56 1999
From: jam at quark.emich.edu (Jeff)
Date: Tue, 16 Nov 1999 07:23:56 -0500
Subject: Alright, I give up...chalk this one up to experience...
In-Reply-To: <383145c0.4445469@212.225.36.99>
References: <383145c0.4445469@212.225.36.99>
Message-ID: <19991116072356.A16111@quark.emich.edu>

On Tue, Nov 16, 1999 at 11:56:28AM +0000, Endgamer wrote:
> Alright, I finally give up. I've been bashing at this bloody idea all
> week...maybe I'm incredibly stupid (probably), but I can't see why,
> when I run this as a script from Windows 95, it doesn't return the
> current date setting on my machine. Could somebody put me out of my
> misery? I promise I'll stop bothering you <g>.
> 
> -------------------------------------------------------
> 
> # Date setting program (chunk 1)
> import os
> 
> olddate=os.popen("date",'r')
> 
> olddate=readdate
> 
> print olddate
> 
> -------------------------------------------------------
> 

I'm pretty sure the issue of setting the date on a win32 machine has been
brought up before in this newsgroup, and I would strongly advise doing a
search on deja.com as well as python.org to see what other resources you can
find.

I would suggest, however, that if you want to *query* the current date and
time setting of the machine that you not use 'popen'.. try the 'time'
module instead:

[~] [7:56pm] [jam at toast-pts/2] % python
Python 1.5.2 (#1, Apr 18 1999, 16:03:16)  [GCC pgcc-2.91.60 19981201 (egcs-1.1.1  on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> import time
>>> print time.ctime.__doc__
ctime(seconds) -> string

Convert a time in seconds since the Epoch to a string in local time.
This is equivalent to asctime(localtime(seconds)).
>>> print time.time.__doc__
time() -> floating point number

Return the current time in seconds since the Epoch.
Fractions of a second may be present if the system clock provides them.
>>> print time.ctime(time.time())
Tue Nov 16 07:23:04 1999

regards,
J
-- 
|| visit gfd <http://quark.emich.edu/>
|| psa member #293 <http://www.python.org/> 
|| New Image Systems & Services, Inc. <http://www.newimage.com/>



From init__self at my-deja.com  Mon Nov  8 11:45:49 1999
From: init__self at my-deja.com (init__self at my-deja.com)
Date: Mon, 08 Nov 1999 16:45:49 GMT
Subject: pythonistas or pythoneers?
References: <07111999.2@sanctum.jae> <382633B1.50993D93@callware.com> <y0jn1spni7c.fsf@vier.idi.ntnu.no>
Message-ID: <806unr$6qp$1@nnrp1.deja.com>

> No way - we are all Pythoneers at the frontiers for computer
> science... ;)

Wouldn't that be a Pyoneer then?


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



From htrd90 at zepler.org  Thu Nov 11 17:39:24 1999
From: htrd90 at zepler.org (Toby Dickenson)
Date: Thu, 11 Nov 1999 22:39:24 +0000
Subject: Running a program as another user on Win32
References: <809ea4$ln$1@nnrp1.deja.com> <e11W3.33036$pf4.31242@news-server.bigpond.net.au>
Message-ID: <wkQrONiRar+JbaAjc0tnRkE7IP8v@4ax.com>

"Mark Hammond" <mhammond at skippinet.com.au> wrote:

>emuller at painewebber.com wrote in message <809ea4$ln$1 at nnrp1.deja.com>...
>>I am writing a bunch of scripts that will need to run as Administrator
>>(Domain Administrator) on Windows NT 4.0 workstation clients that
>>belong to an NT domain. I've tried using LogonUser, but I always get
>>the following: pywintypes.api_error: (1314, 'LogonUser', ' A required
>>privilege is not held by the client.'). I've also tried using LogonUser
>>with the local Administrator account on my NT machine with no luck.
>
>You need to explicitely enable the privilege.

....

LogonUser should enable any priviledges not held by the process as
required. However you do need to grant the priviledge to the user who
will be running your script; "Act as part of the operating system" in
this case.

Yes, even Administrators do not have this priviledge by default. The
System account does, so your existing script should work if it's
running as a service.

hope this helps,


Toby Dickenson



From pinard at iro.umontreal.ca  Tue Nov  2 09:36:06 1999
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 02 Nov 1999 09:36:06 -0500
Subject: Assorted Questions regarding the future of Python...
In-Reply-To: "Fredrik Lundh"'s message of "Tue, 2 Nov 1999 10:47:39 +0100"
References: <6D8A17398E28D3119F860090274DD7DB49833A@pces.cadlab.it> <006201bf2517$50edf1f0$f29b12c2@secret.pythonware.com>
Message-ID: <oqogddnep5.fsf@titan.progiciels-bpi.ca>

"Fredrik Lundh" <fredrik at pythonware.com> writes:

> as far as I know, there are no visible projects working on faster ways
> to execute that [PYC] code, but that doesn't mean that they don't exist ;-)

I'm curious.  Does Python use standard optimisation techniques on that
byte code (CSE, strength reduction, and all that)?  I would be tempted to
guess that the stack machine receives code that pretty closely resemble
the source code.  Would the stack machine be able to express optimisations?
For example, could we copy or exchange arbitrary stack elements?

If yes, maybe one could attempt filtering a `.pyc' through an optimiser, at
least for a long experimental phase, with the hope of having the optimiser
integrated right into Python, one of these later days.

If no, would Guido be favourable at opening the door by adding a few more
stack machine instructions, even if not using them himself, to start with?

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





From hnowak at cuci.nl  Mon Nov  8 17:36:10 1999
From: hnowak at cuci.nl (Hans Nowak)
Date: Mon, 8 Nov 1999 23:36:10 +0100
Subject: Image buttons with Tkinter
In-Reply-To: <38270596.A391995C@parlant.com>
Message-ID: <199911082238.XAA06869@dionysus.fw.cuci.nl>

On 8 Nov 99, at 10:17, Thomas Lane wrote:

> I'm trying to create a button that displays an image instead of text, but
> I don't seem to be having any luck. The button just comes up gray. Here is
> a code snippet of what I'm trying to do:
> 
>   # Put remove.gif on a button (actionbar is a frame for various
> controls)
>   deletebutton = Button(actionbar,image=PhotoImage(file="remove.gif"))
>   deletebutton.place(x=0)

This code worked for me (I threw in a label so the button will be 
clearly visible rather than taking up the whole frame):

#--------------------------
# button-with-gif.py

from Tkinter import *

root = Tk()

image1 = PhotoImage(file='burger.gif')

button = Button(root, image=image1)
button.pack()

Label(root, text="Bwahahaha").pack()

root.mainloop()

#------------------------

> What am I doing wrong here?

Dunno, it looked good to me, although I never use the 'place' thingy.

> Also, is there a place on the net where I can find lots of sample
> Python/Tkinter code? I've read Fredrik Lundh's intro, but I haven't been
> able to find any other code or reference material for Tkinter.

As far as I know, the Tk introduction is about the best you can get 
at the moment. <waiting for Tk books...>

--Hans Nowak (zephyrfalcon at hvision.nl)
Homepage: http://fly.to/zephyrfalcon
You call me a masterless man. You are wrong. I am my own master.



From boud at rempt.xs4all.nl  Tue Nov 30 01:53:29 1999
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 30 Nov 1999 06:53:29 GMT
Subject: XPython (X11)
References: <383E5350.1E07536A@xrce.xerox.com> <slrn83tj53.k7h.bernhard@alpha1.csd.uwm.edu> <81ms2n$126$1@essle.valdyas.org> <slrn840dt7.mov.bernhard@alpha1.csd.uwm.edu>
Message-ID: <81vs99$g8$1@essle.valdyas.org>

Bernhard Reiter <bernhard at alpha1.csd.uwm.edu> wrote:
> On 26 Nov 1999 20:54:47 GMT, Boudewijn Rempt <boud at rempt.xs4all.nl> wrote:
>>Bernhard Reiter <bernhard at alpha1.csd.uwm.edu> wrote:
>>> On Fri, 26 Nov 1999 10:30:56 +0100, Emmanuel Pietriga <emmanuel.pietriga at xrce.xerox.com> wrote:

>>> If I want to program a GUI I look in to wxPython.
>>> Or pyGtk or pyFOX or pyFLTK, but XPython... :)

>>Hey! Don't forget PyQt and PyKDE (and PyQt seems to be
>>working under Windows, too, nowadays)...
> Do you have to but the Troll Kit first? >:)

Sorry, I'm a bit late - I've only just now recovered from 
reading the interview with Guido in Linux Journal where he
says he thinks PyKDE/PyQt is no longer developed. Yes, to
compile the stuff for Windows you need Qt for Windows. Of 
course, you also need a compiler, and those seem to cost
money, too.

> And I didn't mention pyGNOME, why should I mention to PyKDE?

Because the two go together - you did mention PyGTK.

> To create an application tied to KDE or GNOME is a mistake in my
> eyes anyway. <wink>

Oh, yes, people will agree with you. I've one regular correspondent
who berates me continually for not writing my apps to run inside
a browser. I don't really care, though - sometimes I need an
application that's integrated with the rest of my work environment,
and sometimes I can do with something standalone.

> people-will-love-it-any-way-ly, yours

They do, they do, too... ;-).

> 	Bernhard

-- 

Boudewijn Rempt  | http://denden.conlang.org



From m.faassen at vet.uu.nl  Mon Nov 15 06:27:37 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 15 Nov 1999 11:27:37 GMT
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <JlFX3.3659$s87.67930@dfiatx1-snr1.gtei.net>
Message-ID: <80oqn9$ocq$2@newshost.accu.uu.nl>

Jules Allen <usenet-mm-yyyy at julesallen.com> wrote:
> In article <3d1z9sq2uc.fsf at amarok.cnri.reston.va.us>, Andrew M. Kuchling wrote:
> | The issue of indexing modules is an old one, and I'm still astounded
> | that we've never converged on a solution, though everyone always
> | complains about this.  

> Thanks for the History lesson! This helps put it all in perspective
> but why *something* hasn't been built before now is kind of weird.
> The best part of your story for me is confirmation that open source
> projects don't do well without strong leadership. No ESR or an
> ESR-like figure to crack the whip, no Trove.

Actually, 'strong leadership' is not enough. If someone cracks his whip
impressively in open source but doesn't write any code himself, you won't
get any software in the end. So you need someone who actually takes on
the responsibility for writing the code. :)

[Vaults of Parnassus]

Browses through them eagerly and curiously.  :)

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?



From bjorn at roguewave.com  Mon Nov 29 16:49:24 1999
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Mon, 29 Nov 1999 14:49:24 -0700
Subject: Htmlizing text
Message-ID: <51C9F3C392D0D111A15600A0C99E64A9AF569F@localhost>

> Is there a function in the standard Python library to HTML-ize text,
> i.e. to replace 'a > b & c < d' with 'a &gt; b &amp; c &lt; d'?

This is the one I use (look for text2html below).  It does a little bit more
than what you asked for, like turning URLs into clickable links, changing
*word* to italicized etc.

(and yes, it's extracted almost verbatim from the faqwizard :-)

-- bjorn

import string
import re

def translate(text, pre=0):
    translate_prog = prog =
re.compile(r'\b(http|ftp|https)://\S+(\b|/)|\b[-.\w]+@[-.\w]+')
    i = 0
    list = []
    while 1:
        m = prog.search(text, i)
        if not m:
            break
        j = m.start()
        list.append(escape(text[i:j]))
        i = j
        url = m.group(0)
        while url[-1] in '();:,.?\'"<>':
            url = url[:-1]
        i = i + len(url)
        url = escape(url)
        if not pre:
            if ':' in url:
                repl = '<A HREF="%s">%s</A>' % (url, url)
            else:
                repl = '<A HREF="mailto:%s">&lt;%s&gt;</A>' % (url, url)
        else:
            repl = url
        list.append(repl)
    j = len(text)
    list.append(escape(text[i:j]))
    return string.join(list, '')

def escape(s):
    s = string.replace(s, '&', '&amp;')
    s = string.replace(s, '<', '&lt;')
    s = string.replace(s, '>', '&gt;')
    return s

def escapeq(s):
    s = escape(s)
    s = string.replace(s, '"', '&quot;')
    return s
    
def emphasize(line):
    return re.sub(r'\*([a-zA-Z]+)\*', r'<I>\1</I>', line)
    
def text2html(body):
    res = []
    pre = 0
    raw = 0
    for line in string.split(body, '\n'):
        tag = string.lower(string.rstrip(line))
        if tag == '<html>':
            raw = 1
            continue
        if tag == '</html>':
            raw = 0
            continue
        if raw:
            res.append(line)
            continue
        if not string.strip(line):
            if pre:
                res.append('</PRE>')
                pre = 0
            else:
                res.append('<P>')
        else:
            if line[0] not in string.whitespace:
                if pre:
                    res.append('</PRE>')
                    pre = 0
            else:
                if not pre:
                    res.append('<PRE>')
                    pre = 1
            if '/' in line or '@' in line:
                line = translate(line, pre)
            elif '<' in line or '&' in line:
                line = escape(line)
            if not pre and '*' in line:
                line = emphasize(line)
            res.append(line)
    if pre:
        res.append('</PRE>')
        pre = 0
    return string.join(res)



From fredrik at pythonware.com  Fri Nov 12 10:06:38 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 12 Nov 1999 16:06:38 +0100
Subject: Tkinter PhotoImage Question
References: <382C29D7.DBA3E66@callware.com>
Message-ID: <009801bf2d1f$85953ed0$f29b12c2@secret.pythonware.com>

Ivan Van Laningham <ivanlan at callware.com> wrote:
> Ok, I've managed to create an image, using xx=PhotoImage(file), and I
> can retrieve pixel values from that image using xx.get(x,y).
> 
> Now I want to use the put() method to write a pixel.
> 
> How do I do that?  I would have expected that put() would take a string
> of pixel values for the data argument, but that doesn't seem to be the
> case. ...

from the eff-bot archives:

#
# create a greyscale ramp using pure Tkinter
#
# fredrik lundh, january 1998
#
# fredrik at pythonware.com
# http://www.pythonware.com
#

from Tkinter import *

# must initialize interpreter before you can create an image
root = Tk()

data = range(256) # 0..255

im = PhotoImage(width=len(data), height=1)

# tkinter wants a list of pixel lists, where each item is
# a tk colour specification (e.g. "#120432").  map the
# data to a list of strings, and convert the list to the
# appropriate tuple structure

im.put( (tuple(map(lambda v: "#%02x%02x%02x" % (v, v, v), data)),) )

# resize the image to the appropriate height

im = im.zoom(1, 32)

# and display it

w = Label(root, image=im, bd=0)
w.pack()

mainloop()

# after playing with this a little, you'll love
# the Python Imaging Library.




From ivanlan at callware.com  Mon Nov  8 10:33:04 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Mon, 08 Nov 1999 08:33:04 -0700
Subject: Graphics.
References: <3826E949.BBA1BAFD@imf.au.dk>
Message-ID: <3826ED30.70FBE3D3@callware.com>

Hi All, Emil--

Emil Hedevang Lohse wrote:
> 
> How do I draw some graphics on a screen using Python running on
> HP-UX
> MacOS
> Windows
> 

Use the Tkinter Canvas widget.  Start at

	http://www.pythonware.com

& check out the docs.

<seen-one-graphic-ya-seen-em-all>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From tim_one at email.msn.com  Wed Nov 17 03:56:35 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 17 Nov 1999 03:56:35 -0500
Subject: docs
In-Reply-To: <14385.28690.800102.134431@weyr.cnri.reston.va.us>
Message-ID: <000501bf30d9$a6e93100$a42d153f@tim>

[Tim]
> Ask MS for the source code <wink>.  Short of that, there's no reason to
> suppose it's a problem with the docs, and every reason to suspect the
> obvious.

[Fred L. Drake, Jr.]
>   The catch is that no matter how obvious it is, I don't want to lock
> people out of the docs by requiring they update to a newer service
> pack; ...

No, I doubt the service pack has anything to do with it.  What's "obvious"
in the MS world is that it's a random bug that maybe 1 person in 1000 can
reproduce.  That's par for the MS course!  Or did many people report the
same problem with NT SP4?  I saw one report here, and a partial possible
kinda confirmation <wink>.  I certainly never had this problem under any
flavor of NT or of IE.

The usual MS fix is to scandisk, defrag, clear all the caches, and reinstall
stuff from scratch.  And I haven't had more than 8 browser bugs that weren't
eventually cured by reformatting my hard drives <wink>.

otoh-my-sister-has-never-gotten-netscape-to-work-right-ly y'rs  - tim





From mwh21 at cam.ac.uk  Mon Nov  1 08:35:53 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 01 Nov 1999 13:35:53 +0000
Subject: How can I make my assertions smarter?
References: <381D7087.BF8510E5@webone.com.au>
Message-ID: <m34sf6e3ly.fsf@atrus.jesus.cam.ac.uk>

Stuart Hungerford <stuart.hungerford at webone.com.au> writes:

> Hi all,
> 
> I'm using some simple functions to implement crude precondition,
> postcondition and invariant checking in Python classes.
> 
> (Tim: please be gentle ;-)
> 
> Is there some neat way to wrap these tests up so I can print any
> code expression that fails, e.g:
> 
>      Precondition failed foo.py line 42: spoon.shape == 'bent'
> 
> which would appear at line 42 of foo.py as
> 
>      pre(spoon.shape == 'bent')

Hmm... tricky.

Something like this might work (tho' not with python -O, not that that
should be a problem).

chuck it in a file called "pre.py". It will have problems with line
continutations & stuff, but I did just make it up on the spot (well,
about half an hour of spot; must now do something of some use).

class PreconditionViolation(AssertionError):
    def __init__(self,why):
        AssertionError.__init__(self,why)

def pre(cond):
    if cond:
        return
    try: 
        raise ""
    except:
        import sys
        frame = sys.exc_traceback.tb_frame.f_back
        lineno = frame.f_lineno
        filename = frame.f_code.co_filename    
    try:
        import re
        file = open(filename)
        for i in range(lineno):
            line = file.readline()
        match = re.match("\W*(pre\.)?pre\((.*)\);?\W*$",line)
        if match is not None:
            line = match.group(2)
        else:
            import string
            line = string.strip(line)
    except:
        line = "<not found>"
    line = "line %d of %s: %s"%(lineno,filename,line)
    raise PreconditionViolation, line
    
def checked_sqrt(x):
    from math import sqrt
    pre(x>=0)
    return sqrt(x)
    
if __name__=='__main__':
    print checked_sqrt(2)
    print checked_sqrt(-1)


then `python pre.py' gives:

1.41421356237
Traceback (innermost last):
  File "pre.py", line 39, in ?
    print checked_sqrt(-1)
  File "pre.py", line 34, in checked_sqrt
    pre(x>=0)
  File "pre.py", line 30, in pre
    raise PreconditionViolation, line
__main__.PreconditionViolation: line 34 of pre.py: x>=0

but, really, what's wrong with asserts?

Another approach would be to poke around with frame.f_lasti and
decompile the bytecode (!).

> I can see how to use environment variables or global variables to
> selectively enable precondtions etc, although any hints for doing that
> would be appreciated too.

That should be pretty straightforward.

> Thanks in advance,
> 
> Stu

HTH,
Michael



From JamesL at Lugoj.Com  Thu Nov 25 00:03:19 1999
From: JamesL at Lugoj.Com (James Logajan)
Date: Wed, 24 Nov 1999 21:03:19 -0800
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <383979F7.25D6@cs.bham.ac.uk>  <81fsab$2b$1@news.udel.edu> <slrn83nrva.7gk.thantos@chancel.org>
Message-ID: <383CC317.21B89F4C@Lugoj.Com>

Alexander Williams wrote:
> 
> On 24 Nov 1999 05:15:55 GMT, Terry Reedy <tjreedy at udel.edu> wrote:
> >You are currently welcome to save 3 keystrokes (rather than 4)
> >by naming the method parameter 's' instead of 'self'.  The latter
> 
> In a fit of lazy genius, I struck upon using 'me' instead; its a more
> active referent /and/ faster to type than 'self.'  :)

But shouldn't you use the possessive form: "my"? Example:

class C:
   def F(my, arg):
      print my.arg

versus:

class C:
   def F(me, arg):
      print me.arg

(The latter comes out sounding like a crusty old sailer!)



From olivier at scripps.edu  Wed Nov 10 16:23:38 1999
From: olivier at scripps.edu (Olivier Roche)
Date: Wed, 10 Nov 1999 13:23:38 -0800
Subject: random function
References: <3829D78B.7FC142F4@scripps.edu> <19991110.21195500@robert.polytopic.com>
Message-ID: <3829E259.C4A9C6A6@scripps.edu>

Robert Meegan wrote:

> Olivier -
>
> The is both a random and a whrandom module included with the Python
> distribution. Both are identified in the Python Library Reference
> index.
>

You are right !!
I am sorry, i have look at the Python Reference Manual index.

Olivier




From tim_one at email.msn.com  Tue Nov 16 01:31:41 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Tue, 16 Nov 1999 01:31:41 -0500
Subject: docs
In-Reply-To: <14380.25301.616025.322479@weyr.cnri.reston.va.us>
Message-ID: <000a01bf2ffc$3e70b0e0$042d153f@tim>

[Ionel Simionescu, having trouble viewing Python docs]
> I have seen it before and I have just checked it again.
>
> Just open a doc,
> say C:\Python\Doc\tut\tut.html,
> with Internet Explorer 5.00.2314.1003,
> on Windows NT 4.0 SP4,
> and refresh it (F5).

[Fred L. Drake, Jr.]
>   I don't see this when using NT4 SP3; can anyone check this for other
> service packs?

No problem with NT4 SP5.  Same IE, but with a long list of patches.

> I'm really not sure how to approach this problem.  Suggestions from
> people more NT-aware than myself?

Ask MS for the source code <wink>.  Short of that, there's no reason to
suppose it's a problem with the docs, and every reason to suspect the
obvious.

ms-is-out-to-crush-python-ly y'rs  - tim





From malcolmt at smart.net.au  Mon Nov  8 22:30:23 1999
From: malcolmt at smart.net.au (Malcolm Tredinnick)
Date: Tue, 9 Nov 1999 14:30:23 +1100
Subject: Can't use len() on returned PyList object?
In-Reply-To: <3.0.6.32.19991108211217.007cb210@mail.in-gen.net>; from sessile@in-gen.net on Mon, Nov 08, 1999 at 09:12:17PM -0500
References: <3.0.6.32.19991108211217.007cb210@mail.in-gen.net>
Message-ID: <19991109143023.A1001@Ridcully.home>

On Mon, Nov 08, 1999 at 09:12:17PM -0500, sessile at in-gen.net wrote:
> Why can't I use 'len()' to find the length of a list
> returned from my new extension module?  The error is
> shown as:
> 
>     TypeError: call of non-function (type int)

You have probably used 'len' as an integer variable at some point
previously in the program. This would have overwritten the builtin
declaration of len as a function.

You can test this in the interpreter by seeing what type(len) returns (or
even just print len). Programmatically, see if type(len) == type(1). If it
does then you (or one of your imported modules) have redefined len. :-(

Tracking this sort of thing down can be annoying sometimes, because not
only do you have to look through your own program, but also be suspicious
of any modules where you did "from module_name import *" (although, of
course, you can pretty much eliminate all the standard modules, since they
have withstood the test of time for this conflict).

Cheers,
Malcolm Tredinnick



From greg.ewing at compaq.com  Mon Nov  8 06:17:19 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Tue, 09 Nov 1999 00:17:19 +1300
Subject: Get parameter by name
References: <Pine.LNX.4.20.9911061213410.18974-100000@fep132.fep.ru>
Message-ID: <3826B13F.1C8DFA24@compaq.com>

Oleg Broytmann wrote:
> 
>    x = "ka"
> 
> how can I get the value of parameter, the name of which is in x ("ka", in
> this case)?

   foo = locals()[x]

Greg



From mal at lemburg.com  Mon Nov 15 03:57:58 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Mon, 15 Nov 1999 09:57:58 +0100
Subject: Production Web Deployment?
References: <199911080604.BAA22580@python.org> <NBBBKMMANMGEEBHPJJECGEGIFJAA.dirkl@home.com> <14376.24392.828292.926057@dolphin.mojam.com> <80o895$ebn$1@nnrp1.deja.com>
Message-ID: <382FCB16.62B9710A@lemburg.com>

damien_morton at my-deja.com wrote:
> 
> Im working at a small publishing outfit, and we've been pumping out 80
> million hits a month for www.maximmag.com, from a 266MHz IIS/ASP box.
> Weve just moved onto a 3-way clustering server, and expect to be able
> to take pretty much anything that can happen. From time to time, our
> site get mentioned on TV or by Howard Stern and it goes berserk, but we
> expect to be able to handle that too.
> 
> ...
>
> Ive only tried Python within an obscure branch of WSH called SSH,
> without success. I havent tried it as a scripting langauge replacement
> for VBscript, but there has to be a significant advantage to re-
> educating all the vbscript people I work with. For the work they do,
> VBscript is functional and they never seem to complain about it
> constraining them.

AFAIK, you can use Python as ASP language too, even side-by-side
with VBScript. It may well be worth a try...

Just for reference, I am working on a Python based web application
server which can handle around 20 requests/second on my old
AMD K6/266 machine. That's around 1.7 million requests per day
or 51 million per month. Take it to a fast dual processor engine and
it should pretty much cover your needs using a single machine.

Note that it is not Python slowing down these servers, it is
usually one or more of these:
(a) poor socket implementations
(b) poor database connectivity
(c) poor overall design

-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                    46 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/




From m.faassen at vet.uu.nl  Thu Nov 18 05:20:40 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 18 Nov 1999 10:20:40 GMT
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <JlFX3.3659$s87.67930@dfiatx1-snr1.gtei.net> <80oqn9$ocq$2@newshost.accu.uu.nl> <oLUX3.286$yA4.4821@dfiatx1-snr1.gtei.net> <80p9su$ppo$2@newshost.accu.uu.nl> <80piru$13rl$1@hub.org> <80re5d$55a$1@newshost.accu.uu.nl> <fjPY3.770$An3.14793@newsfeed.avtel.net>
Message-ID: <810jto$lo0$1@newshost.accu.uu.nl>

Phil Mayes <nospam at bitbucket.com> wrote:
> Martijn Faassen wrote in message <80re5d$55a$1 at newshost.accu.uu.nl>...
>>T. wrote:
>>> On 15 Nov 1999 15:46:38 GMT, Martijn Faassen <m.faassen at vet.uu.nl> wrote:
>>[snip]
>>
>>You're hard to keep track of too, your listed email address keeps changing!

> Ah, but only the vowels, so we can deduce that it took no more than 6**3 =
> 216 people to develop this.

No, no, a more extensive deja search indicated to me that this person has
posted under other names to the group previously. At least, unless there are
10 people with the same speech patterns posting to comp.lang.python. :)

Anyway, it just made me curious.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?



From paul.robinson at quantisci.co.uk  Tue Nov 30 12:59:54 1999
From: paul.robinson at quantisci.co.uk (Paul Robinson)
Date: Tue, 30 Nov 1999 17:59:54 +0000
Subject: ? on working with http redirects in htmllib
References: <821146$266$1@ash.prod.itd.earthlink.net>
Message-ID: <3844109A.789EDC7E@quantisci.co.uk>

Levi Cook wrote:
> 
> Does anyone know how to process and follow a redirect in python?
> 
> I can easily tell if I've recieved the 301 or 302 errcode, but I cannot for
> the life of me figure out how to get the URL of the new destination.

The URL is simply in the "Location" header.

so for example (untested code):

def actualGet(server, uri):
        got = 0
        while not got:
            h = httplib.HTTP(server)
            h.putrequest('GET', uri)
            h.putheader('Accept', 'text/html')
            h.endheaders()
            errcode, errmsg, headers = h.getreply()
            if errcode == 200:
                f = h.getfile()
                got = 1
            elif errcode == 302 or errcode == 301:
                url = headers.getheader('location')
                #Split into uri and server
                URL2Fetch = urlparse.urlparse(address)
                server = URL2Fetch[1]
                uri = URL2Fetch[2]
            elif errcode == 404:
		#Do something else!
                got = 1
            else:
                raise RuntimeError, ('Unsupported errcode', errcode,
errmsg)
        return f

Hope this helps,
	Paul.



From rdu105 at psu.edu  Wed Nov 10 03:21:52 1999
From: rdu105 at psu.edu (Rhys Ulerich)
Date: Wed, 10 Nov 1999 03:21:52 -0500
Subject: Problem Installing PIL (module search path problem I think)
Message-ID: <38292B20.88C82B0@psu.edu>

Dumb question I'm sure, but I'm fairly stuck so here goes:

I'm trying to install the Python Image Library (PIL) v1.0
Final on my system (mainly for the sake of using the vector
drawing program sketch).


I've successfully compiled PIL in Python-1.5.2/Imaging-1.0.
The installation instructions provided state that the module
must then be added to a location where Python can find the
PIL library.

I'm going to reproduce some of the install instructions here so
please humor the length of this post...

---8<----
[from within Python-1.5.2/Imaging-1.0/]
8. To build a dynamically loaded module, just type "make":
 
        $ make
 
   This will create a file called "_imaging.so".
 
   Then type (assuming a standard shell):
 
        $ PYTHONPATH=.:./PIL ; export PYTHONPATH
        $ python
 
        >>> import _imaging
        >>> import Image
 
   If both imports works, you've successfully added PIL to your Python
   environment.

9. Copy the "_imaging.so" module and the contents of the "Imaging/PIL"
   directory to a suitable place along the Python path.
 
   If you're using Python 1.5 or later, the preferred way is to create
   a "PIL" subdirectory under "site-packages", copy the "PIL.pth" file
   to "site-packages", and the rest of the files to the new 	    
   subdirectory.
 
   Example:
 
    $ cp PIL.pth /usr/local/lib/Python1.5/site-packages
    $ mkdir /usr/local/lib/Python1.5/site-packages/PIL
    $ cp _imaging.so PIL/* /usr/local/lib/Python1.5/site-packages/PIL

----8<----

Step 8 listed above works like a charm.  Step 9 is what is giving me
problems.  I couldn't find a Python-1.5.2/site-packages directory so
I created one and then followed the example commands in step 9 to
put PIL.pth and the _imaging.so library into Python-1.5.2/site-packages.
No luck there.

So I edited the PYTHONPATH environmental variable so that
"PYTHONPATH=.:/usr/local/src/Python-1.5.2/site-packages" but
I still had no luck verifying that "import _imaging" and "import
Image" would work as they did in installation step 8 above.


Any ideas?  I'm sure that this is a really simply thing and I'm
just not RTCFM (the "correct" manual case of an old favorite).

Thank you for your time,
Rhys Ulerich



From rv at eng.fore.com  Wed Nov 24 11:44:20 1999
From: rv at eng.fore.com (Rajesh Vaidheeswarran)
Date: 24 Nov 1999 11:44:20 -0500
Subject: whitespace.py module
Message-ID: <y7b66yrdel7.fsf@agastya.eng.fore.com>

Hello

I'm new to the python language, and the newsgroup. So, if this newsgroup is
not appropriate for this type of posting, please pardon my ignorance and
notify me.

I have written up a module called `whitespace.py' available at
<URL:http://www.dsmit.com/python/>

and a sample program `whitespace' which uses this module, also available at
the same URL.

The module can be used to check a given file for whitespace problems as
defined in the module's documentation (also available as a separate text
document at <URL:http://www.dsmit.com/lisp/whitespace.txt>.

Please feel free use these, and also to send in your comments on either of
these.

I would like to know how to submit a module for inclusion into the standard
python distribution, and would appreciate any pointers in that direction.

Thanks
rv



From I.Clarke at strs.co.uk  Tue Nov  2 05:56:43 1999
From: I.Clarke at strs.co.uk (Ian Clarke)
Date: Tue, 02 Nov 1999 10:56:43 +0000
Subject: Comparison between Python and "Ruby"
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp> <7vjtrc$hmm$1@newshost.accu.uu.nl>
Message-ID: <381EC36B.F479CA1E@strs.co.uk>

> Agreed. This isn't a flamewar anyway. Ian Clarke chose a far more civilized
> title to his post than he did a while back in comp.lang.perl. ;)

Ugh, me civilized?  Now there is flamebait if I ever saw it ;-)

Ian.

--
Ian Clarke                        Email: I.Clarke at strs.co.uk
Homepage: http://www.gnu.demon.co.uk/
Also see: http://freenet.on.openprojects.net/
"All we see and all we seem is but a dream within a dream"



From wjk at wjk.mv.com  Sun Nov 28 22:09:43 1999
From: wjk at wjk.mv.com (William J. King)
Date: Sun, 28 Nov 1999 22:09:43 -0500
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <81bmmi$279$1@nnrp1.deja.com>
Message-ID: <3841EE77.6B5D6AC9@wjk.mv.com>

Python is very interesting and I do not have any real complaints since most

things can be worked around in one way or another. As a relatively new
person
to python programming I have not really compiled a list of complaints:

1.  like the indents -- makes coding easier to read and follow  -
        requires discipline me thinks

2. would like something like  'increment'/'decrement' operators
        either in form:
                    "x++"  or "incr x" vice "x = x + 1 " but its ok

3. would like to be able to iterate through a list using 2/more variables
instead of
        only one (unless its possible and I missed it) :
                            for (x,y,z) in list:
                                    pass            #favorite thing about
Python

4. conversion of integer to a string ( which may exist but havn't found
yet)
        so I made a dictionary for what I needed -- was difficult to
subscript
        strings by adding a number to it without it

5. would like to see regular array construct in addition to
dictionary/list/tuples
        faked that one with a dictionary

6. some simpler description of current regular expressions in python
        some knowledge of grep/vi/awk/sed/perl/tcl - glob - regexp & regsub
and
        now python can be a dangerous thing -- regular expressions rule if
you
        get them right

7. Tkinter is proving to be easier than I thought -- mostly intuitive from
        some previouse experience with Tcl/Tk - X-windows/Win95

8. OOP is different and strange -- new to OOP -- have been looking for
        simple directions in that - Python has certainly helped in making
me
        aware of OOP and providing a better understanding than some
        materials I have read... looked at Java and threw up my hands for
the
        time being

9. Scoping is a bit different that I thought -- try to rember LGB

OK -- now these are not really complaints but I would iterate that I
feel Python is fun and the Python community is great! --- havn't been
spammed for any of my newbie questions and everyone seems to be
very helpful.

I have done a lot of the things you can do in Python by using just
shell scripting and UNIX utilities alone,  so I don't really have any
complaints about Python.






From fredrik at pythonware.com  Tue Nov  2 03:41:41 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 2 Nov 1999 09:41:41 +0100
Subject: Doing network traffic in the 'background' from Tkinter app??
References: <kspuxy62ub.fsf@telelogic.se> <19991029085829.A106971705@vislab.epa.gov> <ksu2n95191.fsf@telelogic.se>
Message-ID: <00b901bf250e$16a33db0$f29b12c2@secret.pythonware.com>

(Andrew, maybe you should fix your reply-to?)

Andrew Markebo wrote:
> | I can't tell exactly what you're doing from your description, but check
> | out Tk's fileevent callback:
> 
> Bingo, Thanx!
> 
> Basically what I want is to give the user the possibilty to mess
> around with my application while it fetches data from THE internet(TM)
> 
> Darn, it seems to be Unix only... fungus.. I would like my application
> to run on windows also..

forget about fileevents, and use Python's own
asynchronous communications framework in-
stead:

http://www.python.org/doc/current/lib/module-asyncore.html

search the newsgroup archives for more information
on using asyncore with tkinter (basically, you should
keep calling asyncore.poll from an idle task).

for more asyncore examples, check the Medusa
servers (http://www.nightmare.com).  you'll also
find a bunch of asyncore/asynchat examples in
the eff-bot guide.

</F>

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




From sekter at mail.matav.hu  Fri Nov  5 11:07:05 1999
From: sekter at mail.matav.hu (Arpad Kiss)
Date: Fri, 5 Nov 1999 17:07:05 +0100
Subject: finding out console user (Win95)
References: <941814380.235290@cdwork.cvt.stuba.sk>
Message-ID: <001501bf27a7$ce30ce70$029f38c3@geometria.hu>

I don't know what "console user" means, but win32api.GetUserName() returns
the current user name.
Arpad

----- Original Message -----
From: Radovan Garabik <garabik at melkor.dnp.fmph.uniba.sk.spam>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Friday, November 05, 1999 3:07 PM
Subject: finding out console user (Win95)


>
> I am porting my script from unix to windows, and I need to find
> console user under win95 (under WinNT I can use os.environ['USERNAME'],
> which is not exactly console user, but good enough for me).
> Something like cuserid() or getlogin() as implemented in
> Cygwin would be ideal....
> Is there a simple way to find it?
>
> TIA





From sjoerd at oratrix.nl  Mon Nov 29 07:04:43 1999
From: sjoerd at oratrix.nl (Sjoerd Mullender)
Date: Mon, 29 Nov 1999 13:04:43 +0100
Subject: XPython (X11)
In-Reply-To: Your message of Fri, 26 Nov 1999 10:30:56 +0100.
             <383E5350.1E07536A@xrce.xerox.com> 
References: <383E5350.1E07536A@xrce.xerox.com> 
Message-ID: <19991129120444.BFD26301CF9@bireme.oratrix.nl>

I use it on a daily basis.  But then, I am responsible for it.
I'll try to help if you need help.

On Fri, Nov 26 1999 Emmanuel Pietriga wrote:

> Is anyone using XPython around here?
> 
> I need help, or at least links to examples (I haven't found any but the
> ones within the package)
> 
> Thanks.
> Emmanuel
> 

-- Sjoerd Mullender <sjoerd.mullender at oratrix.com>



From shouldbe at message.com  Wed Nov 10 12:44:37 1999
From: shouldbe at message.com (Arinte)
Date: Wed, 10 Nov 1999 12:44:37 -0500
Subject: statics everywhere.  Why?
Message-ID: <80cam4$phe$1@rtpnews.raleigh.ibm.com>

Embedding python in a c application, why do I need to make the functions
that python call and variables for callbacks static?

thanks.





From fredp at mygale.org.nospam  Fri Nov 26 09:13:36 1999
From: fredp at mygale.org.nospam (Fred Pacquier)
Date: 26 Nov 1999 14:13:36 GMT
Subject: Running apps with Zope 2?
References: <80v5g8$5a6$1@mulga.cs.mu.OZ.AU> <3834101A.FF39F64C@infercor.no> <8E83A8056PaCmAnRDLM@194.2.0.33> <slrn83jits.cj8.quinn@seniti.ugcs.caltech.edu> <383E7D7A.1B48C7F1@gmd.de>
Message-ID: <8E8A95479PaCmAnRDLM@194.2.0.33>

kanzow at gmd.de (Peter K.) said :

>I have written some web applications using the old Zope/Bobo version. I
>published them using ZopeHTTPServer, like this:
(...)
>After starting ZopeHTTPServer (on hostname.domain.de) with
>ZopeHTTPServer -p 1234 m.py
>I can access the Application object from the web:
>requesting  http://hostname.domain.de:1234/app/hello
>gets "Hello World from my Zope Object".
>This is what I want! But what must I do to achieve this in Zope 2? I
>installed it recently and can't find it in the documentation. How can I
>publish my old applications with the new Zope? Where do I have to put
>m.py? What do I have to do in the management screen? Which URL will get
>me the "Hello World..." message? Do I have to install m.py as a Product?
>How? 

I would love to know the answer too. I have an interesting script I would 
like to run that depends on pcgi and Bobo -- I found pcgi but Bobo seems to 
have disappeared from the surface of the Earth, even Digicool cannot 
provide it anymore. If I could stick this script inside Zope with some 
minimal glue code as you describe, it would solve my problem...

Good-to-know-I'm-not-alone-ly yrs :)

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



From fdrake at acm.org  Mon Nov 29 17:32:48 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Mon, 29 Nov 1999 17:32:48 -0500 (EST)
Subject: python test frameworks
In-Reply-To: <slrn83u3um.4ge.quinn@groat.ugcs.caltech.edu>
References: <slrn83u3um.4ge.quinn@groat.ugcs.caltech.edu>
Message-ID: <14402.65296.413573.98151@weyr.cnri.reston.va.us>

Quinn Dunkan writes:
 > I know this sort of thing has been talked about before, but I'm looking for
 > some information on techniques for self-testing pyhon code (something I
 > should have been doing before, but only now am I writing a system complex
 > enough to warrant it).  I remember Tim's thread about his doc-string embedded
 > testing, and that looked interesting, but I can't find it now.  And links to
 > discussions of the merits of different strategies would be good, too.

  There's also the somewhat-general (and completely undocumented)
"test" package in the Python standard library.  That might be worth a
look; I'm not sure how re-targetable it is.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From JamesL at Lugoj.Com  Thu Nov 25 19:38:54 1999
From: JamesL at Lugoj.Com (James Logajan)
Date: Thu, 25 Nov 1999 16:38:54 -0800
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <867lj85g4c.fsf@g.local>  		<383CC835.B585EBDD@Lugoj.Com> <86r9he2tiz.fsf@g.local>
Message-ID: <383DD69E.5A927386@Lugoj.Com>

Gareth McCaughan wrote:
> 
> James Logajan wrote:
> 
> [I said:]
> >>   - if/then/else works on *statements* (strictly, suites of
> >>     statements) rather than on *expressions*, so I can't
> >>     say
> >>       print "You scored %s point%s" % (score, if score==1 then "" else "s")
> >>     or anything like that. (Or, as some people prefer to put it,
> >>     Python has no ? : operator.)
> >
> > Thanks to Python's dynamic typing you can easily add the C "?:" expression
> > to the language by defining a single function that will work with all types:
> >
> > def ifexp(cond, trueVal, falseVal):
> >    if cond:
> >       return trueVal
> >    else:
> >       return falseVal
> >
> > print "You scored %s point%s" % (score, ifexp(score == 1, "", "s"))
> 
> That has the wrong semantics in the presence of side effects.
> (It's fine for my toy example, but then so is "score<>1 and 's' or ''".

Fair enough. Your requirements weren't just "kind of like C's '?:'
operator," they were precisely that requirement. I'm wondering though what
great application remains unwritten because you are forced to use a
conditional statement rather than use an operator?



From mhammond at skippinet.com.au  Thu Nov  4 16:47:05 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Fri, 5 Nov 1999 08:47:05 +1100
Subject: Assorted Questions regarding the future of Python...
In-Reply-To: <3822f253.11537059@news.demon.co.uk>
Message-ID: <002d01bf270e$238223a0$0501a8c0@bobcat>

[Andy writes:]
> I haven't done any benchmarking myself, but when I installed VB6 I
> didn't exactly leap off my seat at the speed increase.

And more to the point, VB was (and is), usually, "fast enough".  Who
does huge matrix manipulation etc in VB anyway?  Lots of the focus the
industry has on speed is a red herring IMO.

Just like Python itself - really quite "slow" compared to these
natively compiled languages - but in the entire time I have used it, I
have had to use the profiler exactly twice.  In every other case, it
has been more than "fast enough" (and the development speed far
outweighs the difference anyway).

Whats-a-milli-second-between-friends-ly,

Mark.




From mwh21 at cam.ac.uk  Wed Nov 24 10:40:31 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 24 Nov 1999 15:40:31 +0000
Subject: CRC-module
References: <38396766.B189D15A@bibsyst.no> <383C1089.1D0F7C18@bibsyst.no>
Message-ID: <m3zow3sxsg.fsf@atrus.jesus.cam.ac.uk>

Thomas Weholt <thomas at bibsyst.no> writes:

> Hi,
> 
> Ok, so I`ve looked into zlib.crc32 and zlib.adler32. They seem easy
> enough to use, but I thought crc-codes had characters and numbers in
> them, not just a plain integer like the methods above return. ( As you
> can see, I`m a complete ass on this subject, but don`t have time to do
> the proper research myself, and was hoping for a "quick fix" ... )

Well, on one level there's not much difference between a binary string
and an integer. But crc32 returns a 32-bit value, so it's most
convenient/efficient to store it in an integer.
 
> A friend of mine mentioned that I should try SHA-1 instead, for more
> accuracy. Can anybody give me an example on how to compute crc-codes,
> using zlib or preferrably some more accurate method, for single files ??

Well, comparing crc32 and SHA-1 or md5 isn't really comparing like
with like, to the (small) extent of my knowledge on the matter; crc32
(AFAIK) is designed to spot accidental transmission errors, sha-1/md5
are (certainly) designed to spot malicious modification.

Also md5 is 128 bits and sha-1 is 160, so obviously these are finer
grained than crc32.

> If this is all it takes :
> 
> crc = module_name.crc_method(file)
> 
> and comparison is done like :
> 
> if (crc1 == crc2): print "Equal."
> else : print "Different."
> 
> then all I need is the name of the most effective/accurate module to
> use.
> 
> If, for some strange reason, I should use one module instead of another,
> that info would be interesting too.

For that application, it'd probably be best to use sha, eg:

import sha
sha.sha(open(filename).read()).digest()

Efficiency-wise that'll be IO bound so the fact that SHA-1 is more
CPU-intensive (I think) than crc32 shouldn't be relavent.

HTH,
Michael



From uhuenemoerder at hotmail.com  Sat Nov  6 15:06:46 1999
From: uhuenemoerder at hotmail.com (Ulrich Huenemoerder)
Date: Sat, 6 Nov 1999 21:06:46 +0100
Subject: PyRun_File / PyRun_String
Message-ID: <801vbn$2l64$1@bnews.gigabell.net>

Hello,

this might have been asked before, but be sure I looked up deja.com etc.
before I sent this mail.

I would like to embed Python into my C++ Application. I already looked up
the functions PyRun_SimpleFile and PyRun_SimpleString, but since I need
multiple independent interpreters at once I think PyRun_File and
PyRun_String are the functions I need.

Could anyone provide me with the code for a small example app, like

// initialization
Py_Initialize();
...

// initialize the interpreter
...

// parse a file
PyRun_File(...);

// parse a string
PyRun_String(...);

// that's it.

Thanks,

Ulrich.





From mhammond at skippinet.com.au  Wed Nov 24 17:09:06 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 25 Nov 1999 09:09:06 +1100
Subject: Python and COM: a bug?
In-Reply-To: <816010E2456BD111A48700805FBBE2EEFDF822@ex-quebec-u1.baan.com>
Message-ID: <004e01bf36c8$872b1550$0501a8c0@bobcat>

> OK. You may well be right about a bug in our code. I was told that
the
> interface is not dual and that the dispids are created
> dynamically the first
> time a property is used. So maybe there is a problem in this
> code and that
> the fact that I have never encountered the problem in
> VBScript is pure luck.
> I'm still not sure, because a lot of people are using those
> COM objects in
> dozens of VB Scripts and none have reported this kind of
> problem. But, hey,
> I have seen much stranger occurences than that already :)

No luck involved - it could well be that VBScript is simply not
exploiting that COM rule - I will check.

Yes - VBScript does indeed re-query for the DISPID before each call -
ie, it does not cache the items.  I would put money on the fact that
this is your problem.

If it is the problem, you are clearly breaking the rules, and there is
a good chance that later versions of VBScript, or indeed any other COM
compatible language would also break.

I am quite confident that we do not have anything this fundamental
broken.


> What I'll do is write a test script in Python that just tries
> to reproduce
> the problem. If I can finally find a way to reproduce it on
> demand, I'll ask

Or you could take another tact - something like:

ob = pythoncom.CoCreateInstance("Your.ProgID", None,
pythoncom.CLSCTX_ALL, pythoncom.IID_IDispatch)

dispid = ob.GetIDsOfNames("foo")
print dispid

dispid2 = ob.GetIDsOfNames("foo")
print dispid2

if dispid != dispid2:
  print "Eeek"

> That's something I don't know about. My (naive) question is:
> why doesn't
> VBScript require such a step?

Python also doesnt _need_ this step - just does a better job with
byref params etc if you use it.  I doubt it will have any effect on
your problem.

Mark.




From ivanlan at callware.com  Sun Nov 14 17:11:35 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Sun, 14 Nov 1999 15:11:35 -0700
Subject: Other ways to disable Wm in Tkinter?
References: <80nabr$q62$1@nnrp1.deja.com>
Message-ID: <382F3397.F8552F29@callware.com>

Hi All--

brian at liaone.ne.mediaone.net wrote:
> 

[bobbitt]

> Thus, my question - can one disable the window manager (or at least
> disable window movement) short of withdraw/overrideredirect/deiconify?
> It goes without saying that using grab_set_global is a bit too brutal.
> Thanks,
> Brian
> 

Nuking or disabling people's window managers is not as brutal as
grab_set_global?????  <.9 wince>

Why do you need to do this?  I'd advise against it.  I think all you're
going to accomplish is pissing people off.

If you absolutely *must*, try mapping an invisible window that covers
the whole root window and intercepting all the events, only passing on
the required ones to the dialog.  But I think you'd have to write a C
extension to do this, which wouldn't be portable.  And it seems pretty
kludgy to me. ...

<don't-touch-that-wm>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours
----------------------------------------------



From andy at robanal.demon.co.uk  Sun Nov  7 17:05:40 1999
From: andy at robanal.demon.co.uk (Andy Robinson)
Date: Sun, 07 Nov 1999 22:05:40 GMT
Subject: Can Zope be used as a CVS server?
References: <harms-0511991330570001@192.168.1.35> <14371.11961.82152.303247@dolphin.mojam.com> <harms-0611990947260001@192.168.1.35>
Message-ID: <3828f7ac.36579949@news.demon.co.uk>

harms at mbnet.mb.ca (Stefan Harms) wrote:

>In article <14371.11961.82152.303247 at dolphin.mojam.com>, skip at mojam.com (Skip Montanaro) wrote:
>
>>     Stefan> Could I use Zope for the functionality of a CVS server?
>> 
>> Dunno, and it's not obvious (to me) why you'd want to do that.  Could you
>> motivate your question by providing some rationale for not using CVS itself?
>> 
>> Skip Montanaro | http://www.mojam.com/
>> skip at mojam.com | http://www.musi-cal.com/
>> 847-971-7098   | Python: Programming the way Guido indented...
>
>
>The only reason for using Zope as a CVS server, is that somebody I know finds it difficult to get CVS going as a server, and that the potential participants are not familiar with CVS, but they all have browsers...
>and therefore can learn to use Zope.
>
>Thanks for the reply

I agree and asked about this on the Zope list, but got no reply.

If you work behind firewalls then you need a web interface to CVS, a
web-based bug-tracking system, and a web-based project mailing list.
You can get all three separately but it is a pain in the ****.

If someone made a set of Zope objects which picked up the Zope
permissions and user model, and set up and managed a CVS project
accordingly (assuming CVS is on the server), one box such as Starship
could provide a collaborative working environment for a whole host of
projects, and the administrator could stay sane because Zope lets him
delegate management.

- Andy



From gduzan at gte.com  Fri Nov 12 18:20:02 1999
From: gduzan at gte.com (Gary Duzan)
Date: Fri, 12 Nov 1999 18:20:02 -0500
Subject: AIX, xlC, and C++ Modules
Message-ID: <382CA0A2.BE6536D3@gte.com>

   While working on a port of the new OmniORBpy stuff to AIX, I
came across an interesting issue with python's module loader.
importdl.c uses the load() call on AIX to do the loading, which
does fine with C and a fair amount of C++. However, the C++ module
I was trying to load includes static variables with constructors,
and AIX's load() call doesn't handle them properly. To do so, you
have to go to the C++ side of the house and call loadAndInit(),
which is otherwise identical but does the initialization properly.
I'll include the changes that work for me below. I also tried
reverting to dlopen(), but this seemed not to resolve the
Py_InitModule symbol correctly, leading to a quick core dump.
   The patch enclosed fixes the problem for me. If anyone else
can suggest a better alternative, I'd be glad to hear it.

Setup: AIX 4.2.1, C Set++ (xlC) 3.1.4, --with_thread, CC=xlC_r.

					Gary Duzan
					GTE Laboratories
-------------- next part --------------
--- Python-1.5.2/Python/importdl.c	Wed Jan 27 12:53:10 1999
+++ Python-1.5.2/Python/importdl.c.new	Fri Nov 12 17:50:50 1999
@@ -178,7 +178,12 @@
 #define DYNAMIC_LINK
 #define SHORT_EXT ".so"
 #define LONG_EXT "module.so"
+#ifdef USE_SHLIB
+#include <dlfcn.h>
+#else
 #include <sys/ldr.h>
+#include "/usr/lpp/xlC/include/load.h"
+#endif
 typedef void (*dl_funcptr)();
 #define _DL_FUNCPTR_DEFINED
 static int  aix_getoldmodules(void **);
@@ -458,7 +463,7 @@
 		p = (dl_funcptr) dlsym(handle, funcname);
 	}
 #endif /* USE_SHLIB */
-#ifdef _AIX
+#if defined(_AIX) && !defined(USE_SHLIB)
 	/*
 	-- Invoke load() with L_NOAUTODEFER leaving the imported symbols
 	-- of the shared module unresolved. Thus we have to resolve them
@@ -473,7 +478,8 @@
 		if (!staticmodlistptr)
 			if (aix_getoldmodules(&staticmodlistptr) == -1)
 				return NULL;
-		p = (dl_funcptr) load(pathname, L_NOAUTODEFER, 0);
+		/* p = (dl_funcptr) load(pathname, L_NOAUTODEFER, 0); */
+		p = (dl_funcptr) loadAndInit(pathname, L_NOAUTODEFER, 0);
 		if (p == NULL) {
 			aix_loaderror(pathname);
 			return NULL;
@@ -877,7 +883,7 @@
 }
 
 
-#ifdef _AIX
+#if defined(_AIX) && !defined(USE_SHLIB)
 
 #include <ctype.h>	/*  for isdigit()	  */
 #include <errno.h>	/*  for global errno      */

From hildeb at www.stahl.bau.tu-bs.de  Fri Nov 26 09:19:11 1999
From: hildeb at www.stahl.bau.tu-bs.de (Ralf Hildebrandt)
Date: 26 Nov 1999 14:19:11 GMT
Subject: fcntl/ioctl and O_NONBLOCK
Message-ID: <slrn83t5mv.bhk.hildeb@stahlw06.stahl.bau.tu-bs.de>

Hi!

I'm trying the following unsing Python-1.5.2 on Solaris 2.6: 

... snip ...
program = popen2.Popen3(commandline, self.true, buffersize)
# start program with 64k Buffer

stdin  = program.tochild
stdout = program.fromchild

rv = fcntl.ioctl(stdout.fileno(), FCNTL.O_NONBLOCK, 1)
# rv = fcntl.ioctl(stdout.fileno(), FCNTL.O_NDELAY, 1)
... snap ...

But no matter what I do - I keep getting
IOError: [Errno 22] Invalid argument

What I want to do: I want to use select() to see if there's data to be
read from stdin or stdout and read it ALL (thus I need stdin/stdout to
be non blocking!)

Any ideas?



From lijnzaad at ebi.ac.uk  Thu Nov 25 10:01:19 1999
From: lijnzaad at ebi.ac.uk (Philip Lijnzaad)
Date: 25 Nov 1999 15:01:19 +0000
Subject: automatic "lootjes trekken" [ was Re: automatic "lotjes trekken"]
References: <81cqnv$2ip$1@vvs.superst.iae.nl> <81e3s2$j14$2@newshost.accu.uu.nl> <u7903p71c5.fsf_-_@ebi.ac.uk> <81hbrq$1fh$1@vvs.superst.iae.nl>
Message-ID: <u73dtu7gzk.fsf@ebi.ac.uk>

sorry, I still can't resist, even though it's futile ... 

Carel> Well, the file at home *is* called lootjes-trekken.py, but somehow the
Carel> spellling got corrupted:)

Carel> OTOH, Dutch being a permissive language does allow for both variants!

mm ... does it? Living abroad, I haven't kept up with the new spelling. Or
perhaps it's more of a regional thing? (Incidentally, dialects of computer
languages never seem to be regional. Oh, and (also incidentally) please don't
get the impression that Dutch is indentationally-structured :-)

Carel> "lotjes" being the generic form, "lootjes" the special case (from a
Carel> morphological point of view that is). 

To me, the form 'lotje' resonates more with 'van lotje getikt' (meaning 'mad,
stupid, bonkers' which is an admittedly fairly old-fashioned expression) than
with 'lot' (meaning 'lottery ticket').

Still-not-resisting-the-futile-urge-ly yours,

                                                                      Philip
-- 
If information has state, are we now in the liquid, gaseous or plasmatic phase?
-----------------------------------------------------------------------------
Philip Lijnzaad, lijnzaad at ebi.ac.uk | European Bioinformatics Institute,rm A2-24
+44 (0)1223 49 4639                 | Wellcome Trust Genome Campus, Hinxton
+44 (0)1223 49 4468 (fax)           | Cambridgeshire CB10 1SD,  GREAT BRITAIN
PGP fingerprint: E1 03 BF 80 94 61 B6 FC  50 3D 1F 64 40 75 FB 53



From fredrik at pythonware.com  Thu Nov  4 03:31:46 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 4 Nov 1999 09:31:46 +0100
Subject: opening anydbm files
References: <7vdtet$3i4$1@marcie.fissiontech.com>
Message-ID: <014101bf269f$09019f20$f29b12c2@secret.pythonware.com>

Scott Barron wrote:
> According to the docs and Fredrik's book this should work but I get the
> following:
> 
> Python 1.5.2 (#8, Oct 13 1999, 22:39:47)  [GCC 2.7.2.3] on linux2
> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> >>> db = anydbm.open ("foo", "r")
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
>   File "/usr/local/lib/python1.5/anydbm.py", line 80, in open
>     raise error, "need 'c' or 'n' flag to open new db"
> anydbm.error: need 'c' or 'n' flag to open new db

you get this error if whichdb fails to identify the
existing database -- the most likely reason is that
anydbm("c") used the "dumbdbm" handler.

try running:

    import anydbm
    print anydbm._defaultmod

to see what the default database handler is.

</F>

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




From gerglery at usa.net  Mon Nov  1 06:36:23 1999
From: gerglery at usa.net (greg andruk)
Date: 01 Nov 1999 06:36:23 -0500
Subject: Need Berkley DB!!
References: <381D5E84.98931682@bibsyst.no>
Message-ID: <87d7tuxx3c.fsf@slip-32-100-243-27.ma.us.prserv.net>

Thomas Weholt <thomas at bibsyst.no> wrote:

> After I spent most of my weekend trying to get Berkley DB to work with
> Python on my Linux ( Suse 6.1) machine, I`ve now given up completly.
> I`ve tried to install DB.1.85. and DB.2.7.7, neither worked, either
> crashing under installation ( which of course, is not Python
> related ) or when I tried to test the installation with the examples in
> the documentation ( the wrapper for the 2.7.5-module listed at
> www.python.org, under modules, databases ).

You do have to much with the headers to make the stock db module
compile with DB 2.x.  If 1.85 isn't working for you , you've got,
well, a problem.

Python also supports GNU db, and pretty much every flavor of Linux
comes with that.  Can you use it instead?



From dieter at handshake.de  Fri Nov 12 14:22:08 1999
From: dieter at handshake.de (Dieter Maurer)
Date: 12 Nov 1999 20:22:08 +0100
Subject: Minor Tkinter omission: grid_remove
Message-ID: <x766z7o6qn.fsf@lindm.dm>

Tkinter (1.5.2) does not implement a "grid_remove" method
(for the tk "grid remove slave ..." command).

Workaround:
  Instead of "w.grid_remove()" use "w.tk.call('grid','remove',w._w)"

- Dieter



From kuncej at mail.conservation.state.mo.us  Fri Nov 12 17:31:56 1999
From: kuncej at mail.conservation.state.mo.us (Jeffrey Kunce)
Date: Fri, 12 Nov 1999 16:31:56 -0600
Subject: XML-RPC for CGI
Message-ID: <s82c413c.005@mail.conservation.state.mo.us>

I have a web host where I can't use long-running processes (like 
an XML-RPC server). I can run CGI scripts on the host, and want
to take advantage of the XML-RPC protocol.

So I have written xmlrpccgiserver.py (based on xmlrpcserver.py)
It is included in the attached tarball, along with a demo cgi script 
and demo client script. 

I haven't put it into production yet, but the tests look good.

  --Jeff


-------------- next part --------------
A non-text attachment was scrubbed...
Name: xmlrpccgi.tgz
Type: application/x-compressed
Size: 2122 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/19991112/75f7ad83/attachment.bin>

From bbosware at vic.bigpond.net.au  Tue Nov 30 16:47:05 1999
From: bbosware at vic.bigpond.net.au (John Leach)
Date: Tue, 30 Nov 1999 21:47:05 GMT
Subject: Looking for sshlib.py
References: <iwI04.9621$VA6.47863@news-server.bigpond.net.au> <19991130080326.A1376@quark.emich.edu>
Message-ID: <tBX04.11350$VA6.54975@news-server.bigpond.net.au>

>what type of application are you trying to write that you need this type of
>access?

I'm looking to write a Python frontend for character-based applications
accessed via the internet.
telnet works well but I'd like something with more security (telnet sends
its password unencrypted I believe).
I'll try calling ssh from within Python and see if that helps.
Thanks for your reply,
John Leach
http://www.bosware.com.au







From bwinton at tor.dhs.org  Thu Nov  4 19:49:22 1999
From: bwinton at tor.dhs.org (Blake Winton)
Date: Fri, 05 Nov 1999 00:49:22 GMT
Subject: Snippets site (minor rant)
Message-ID: <slrn824efu.1e6.bwinton@tor.dhs.org>

Well, I just checked the newsgroup, and no-one seems to have noticed
that (or at least no-one has posted anything about ) the Python Snippets
site has been down for the past few days.  I'm not sure if that's a good
or a bad thing.

Anyways.  (Here comes the rant.)  The site was down because Bell (aka
Sympatico, my local ADSL provider) decided to switch their service from
DHCP to Access Manager (aka PPPoE).  The PPPoE implementation for Linux
is "not quite production quality".  So on the day they switched my area
over, I switched to Rogers at Home (my local Cable Modem provider).

The end result is that my computer has just gotten on the net, and the
dns will switch over to it at 6am EST.

I wanted to apologize to Hans, and to vent a little, and now that I've
done both, I'm going to go and help a buddy move into his new apartment.

Thank you all for listening,
Blake.




From ScherBi at BAM.com  Tue Nov 16 14:24:08 1999
From: ScherBi at BAM.com (ScherBi at BAM.com)
Date: Tue, 16 Nov 1999 14:24:08 -0500
Subject: problems with __getattr__ & __setattr__
Message-ID: <3F8BA119B48AD11197470008C7A493CB98B67D@HQOBGEX1>

I am having trouble using __getattr__ and __setattr__ in some classes.  I've
boiled it down to the code below.
It dumps core under 1.5.2 on Linux or WinNT.

I think maybe it's looping, if so, how does one go about doing this? 
(I'm assuming it's clear enough what I'm trying to do.)

If I comment out the __gettattr__ operation, it doesn't dump core.  Instead
I get an AttributeError: eggs raised at the 'self.eggs[key] = value' call in
__setattr__.  This is what leads me to belive there's something circular
going on here.


Thanks,
 Bill

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

class spam:

    def __init__(self):
        self.eggs = {}
    
    def __getattr__(self, key):
        try:
            return self.eggs[key]
        except:
            raise
        
    def __setattr__(self, key, value):
        try:
            self.eggs[key] = value
            return 0
        except:
            raise



if __name__ == '__main__':

    s = spam()
    print 'instance of spam created.' 

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



From edcjones at erols.com  Wed Nov  3 23:40:02 1999
From: edcjones at erols.com (Edward C. Jones)
Date: Wed, 03 Nov 1999 23:40:02 -0500
Subject: Python / Perl joke in rec.humor.funny
Message-ID: <38210E22.BBDAF855@erols.com>

A Python vs. Perl joke has appeared in the moderated,
low-traffic, humor group rec.humor.funny. It can be found at

http://www.netfunny.com/rhf/jokes/99/Nov/perl.html




From fredp at mygale.org.nospam  Mon Nov 15 06:08:36 1999
From: fredp at mygale.org.nospam (Fred Pacquier)
Date: 15 Nov 1999 11:08:36 GMT
Subject: Parnassus
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <JlFX3.3659$s87.67930@dfiatx1-snr1.gtei.net> <382F28A3.A09C3CEF@callware.com>
Message-ID: <8E7F76BFAPaCmAnRDLM@194.2.0.33>

ivanlan at callware.com (Ivan Van Laningham) said :

>Boudewijn Rempt wrote:
>
>> You don't say! I just went there, spurred on by your posting to
>> enter my own efforts. Just when I had done so, I noticed that one
>> of them was already there - really good work! It seems to be very,
>> very nice, indeed.
>
>Same thing happened to me.  I thought, "surely he hasn't added my stuff,
>no one knows about it."  Too late, there they were.  I'm impressed.

Impressive, indeed. I went there to check because I posted (for the first 
time) an announce for a very modest snippet a month or two ago, and sure 
enough, it is there too. It seems considerable effort has been made to seed  
the archive with recent newsgroup content, and that could make a big 
difference... I do hope the effort is rewarded and Parnassus evolves into 
the CPAN or Freshmeat Python users have been lacking !

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



From boncelet at udel.edu  Mon Nov  8 05:09:14 1999
From: boncelet at udel.edu (Charles Boncelet)
Date: Mon, 08 Nov 1999 10:09:14 +0000
Subject: __str__ vs. __repr__
References: <000001bf275c$02370d20$1d2d153f@tim>
Message-ID: <3826A14A.3328AFAA@udel.edu>

Tim Peters wrote:

> David Ascher suggested that str(long) drop the trailing "L", and that's a
> good example.  The "L" is appropriate for repr(), for so long as Python
> maintains such a sharp distinction between ints and longs, but is really of
> no help to most users most of the time.
On the dissenting side, I find the "L" useful (just as I find the
difference between 1 and '1' useful.  Consider the following:

>>> a,b = 2L, 2
>>> a,b
(2L, 2)
>>> a**32
4294967296L
>>> b**32
Traceback (innermost last):
  File "<stdin>", line 1, in ?
OverflowError: integer pow()

Without the "L", it is not at all clear why a**32 works and b**32 fails.

--Charlie

-----
Charles Boncelet <boncelet at udel.edu>
University of Delaware
Newark DE 19716 USA
http://www.eecis.udel.edu/~boncelet/



From reic0024 at bulldog5.d.umn.edu  Tue Nov  2 17:37:34 1999
From: reic0024 at bulldog5.d.umn.edu (Aaron J Reichow)
Date: Tue, 2 Nov 1999 16:37:34 -0600
Subject: How to control app from Python?
In-Reply-To: <381F399E.CC26AE2D@theriver.com>
References: <38160DD8.5C695E09@theriver.com> <381A8A5A.C94398BD@saiph.com> <381F399E.CC26AE2D@theriver.com>
Message-ID: <Pine.SO4.4.02.9911021629330.10075-100000@bulldog5.d.umn.edu>

> I don't know much about the way the app works.  I hope to know a whole
> lot more about it in the next couple of days ;-)
> 
> This is the first time I've tried to write code in any language to
> control another application.  It's very interesting.  Lots to learn!

It's a shame Linux/Unix didn't have some unified API (probably a misuse of
the word) like as in MacOS, where one can access the items in the menus,
&c, with AppleScript.

Aaron




From thomas.heller at ion-tof.com  Thu Nov  4 08:32:46 1999
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Thu, 4 Nov 1999 14:32:46 +0100
Subject: __str__ vs. __repr__
References: <000401bf25b9$88ea57e0$2e2d153f@tim> <199911040157.UAA11908@eric.cnri.reston.va.us> <m31za6339e.fsf@greebo.nodomain.de>
Message-ID: <00c001bf26c9$14f75a20$4500a8c0@thomasnotebook>

> How about a hook function a la __import__? I.e. have a function
> __format__ [1] in __builtin__ that is bound to repr by default. Whenever
> Python prints the result of an expression in interactive mode, it calls
> __format__ with the result as parameter and expects it to return a
> string ready for printing.

But you have a custom python interpreter, don't you? Or how do you
manage to use your __format__ as the printer for interactive mode?

Thomas Heller




From news at dorb.com  Thu Nov 11 13:20:13 1999
From: news at dorb.com (Darrell)
Date: Thu, 11 Nov 1999 13:20:13 -0500
Subject: ASCII delimited files
References: <942231502.1914033961@news.interbusiness.it> <80ekhb$1qrh$1@hub.org> <80et0b$2df8$1@hub.org>
Message-ID: <YMDW3.5675$JD1.379317@typhoon.nyroc.rr.com>

Oops.
That fix pushed into or beyond exponential runtime.
It won't finish the 3.6meg file anytime soon.
String slicing and pasting is dirt slow.

E:\ace\dtd\q\delim1>delimited.py
3600 bytes
0.0310000181198 time

E:\ace\dtd\q\delim1>delimited.py
36000 bytes
3.04700005054 time

Got some improvement with the attached version at
the cost of memory.

Now I'm thinking about the 'C' version again.

E:\ace\dtd\q\delim1>delimited.py
3600
0.0310000181198

E:\ace\dtd\q\delim1>delimited.py
36000
0.327999949455

E:\ace\dtd\q\delim1>delimited.py
360000
3.21899998188

E:\ace\dtd\q\delim1>delimited.py
3600000
48.2030000687

--
--Darrell


begin 666 delimited.py
M(R$@+W5S<B]B:6XO96YV('!Y=&AO;@T*(R!D96QI;6ET960N<'D-"B, at 36]D
M:69I960 at 9G)O;2!T:&4@;W)I9VEN86P at 8GD@1"=!<F-Y($HN32X at 0V%I;B \
M9&%R8WE 9')U:60N;F5T/@T*(PT*(R!#5#DY,#(R-B!6,"XQ#0H-"B(B(@T*
M8G)E86MI;F<@=&%B(&1E;&EM:71E9"!O<B!#4U8@=&5X="!I;G1O(&$@;&ES
M="!O9B!T=7!L97,N#0I/<FEG:6YA;"!V97)S:6]N.B!334, at 9&%T86)O;VL@
M<&QU9VEN+"!R96%D>&QS+G!Y#0I'96YE<F%L:7IE9"!T;R!B92!U<V%B;&4@
M9F]R(&%N>2!D96QI;6ET97(@8G5T(%QN+ at T*(B(B#0H-"FEM<&]R="!S=')I
M;F<-"@T*9&5F(&EN<V5R=$QI<W1$96QE=&4H:6YB=68L(&PI. at T*(" @("(B
M(B!);G-E<G0 at 86YD(&1E;&5T92!S96=M96YT<R!I;B!A(&)U9F9E<@T*(" @
M(&P@:7, at 82!L:7-T(&]F("AS=&%R="P@<V5Q3G5M+"!S=')I;F<L(&5N9"D@
M5&AE(&EN<'5T("=L)R!M=7-T(&)E('-O<G1E9 T*(" @($EF('-T87)T(&%N
M9"!E;F0 at 87)E(&5Q=6%L('1H96X@<W1R:6YG(&ES(&EN<V5R=&5D(&%T('1H
M870@<&]I;G0-"B @("!)9B!E;F0@/B!S=&%R="!T:&5N('1H:7,@<F%N9V4@
M:7, at 9&5L971E9"X-"B @("!)9B!E;F0@/"!S=&%R="!T:&5N('1H:7,@<F%N
M9V4@:7, at 9'5P;&EC871E9 T*#0H@(" @5&AE(&1E;&5T92!M;V1E(&ES(&YO
M="!W96QL('1E<W1E9"X-"B @("!7:&5N(&EN<V5R=&EN9R!W:&EC:"!C86X@
M8F4@=&AO=6=H(&]F(&%S(')E<&QA8V4@:7,@=&AE(&UO<W0 at 8V]M;6]N#0H@
M(" @=7-E+B!#=70 at 9G)O;2!S;VUE('-T87)T:6YG('!O:6YT(&EN(&)U9B!T
M;R!S;VUE(&5N9"!P;VEN="!A;F0-"B @("!I;G-E<G0@<V]M92!N97<@=&5X
M="X-"@T*(" @($%N(&]F9G-E="!I<R!M86EN=&%I;F5D(&-A;&QE9" B;&%S
M="(-"B @("!)9B!A;B!A='1E;7!T(&ES(&UA9&4@=&\@:6YS97)T(&%T(&$@
M<&]I;G0@;&]W97(@=&AA;B!L87-T#0H@(" @:70@:7,@:6=N;W)E9"X at 5&AI
M<R!I<R!T:&4 at 8V%S92!O9B!T=V\@<F5P;&%C97,@=V]R:VEN9R!O;@T*(" @
M('1H92!S86UE('!O:6YT+B!4:&4@<F5P;&%C92!W;W)K:6YG(&%T('1H92!L
M;W=E<W0@:6YD97@@=VEN<RX-"@T*(" @(%1H92!I;G!U="!W;W)K(&QI<W0@
M(FPB(&UU<W0 at 8F4@<V]R=&5D+ at T*(" @("(B(@T*(" @('-P;&ET0G5F/5M=
M#0H@(" @87!P96YD(#T@<W!L:71"=68N87!P96YD#0H@(" @;&%S=#TP#0H@
M(" @:68@;&5N*&PI/3TP. at T*(" @(" @("!R971U<FX@:6YB=68-"B @("!I
M9B!L96XH;%LP72D@/B S. at T*(" @(" @("!S=')/9F8], at T*(" @(" @("!E
M;F1/9F8],PT*(" @(&5L<V4Z#0H@(" @(" @('-T<D]F9CTQ#0H@(" @(" @
M(&5N9$]F9CTR#0H-"B @("!F;W(@:2!I;B!L. at T*(" @(" @("!N>'0]:5LP
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M"@T*9&5F('-P;&ET7V1E;&EM:71E9"AS+"!D96QI;6ET97(]3F]N92D at .@T*
M(" @("(B(G-P;&ET(&$@9&5L:6UI=&5D('1E>'0 at 9FEL92!S=')I;F<@:6YT
M;R!A(&QI<W0@;V8@='5P;&5S+ at T*(" @(%%U;W1E<R!A<F4@;V)E>65D+"!E
M;F-L;W-E9"!N97=L:6YE<R!A<F4 at 97AP86YD960@=&\@=&%B+ T*(" @(&1O
M=6)L92!Q=6]T97, at 9V\@=&\@<75O=&5S(B(B#0H-"B @(" C(#DX,#0R-B!F
M:6YD:6YG(&QI;F4 at 9&5L:6UI=&5R(&1Y;F%M:6-A;&QY#0H-"B @("!P<F]B
M92 ]('-;.C$P,# P70T*(" @(&5O;" ](&9I;F1L:6YE9&5L:6UI=&5R*'!R
M;V)E*0T*#0H@(" @(R Y.3 R,C8 at 9W5E<W-I;F<@9FEE;&0 at 9&5L:6UI=&5R
M(&9R;VT@)UQT+"<@:68@;F]T('-U<'!L:65D#0H-"B @("!I9B!N;W0 at 9&5L
M:6UI=&5R. at T*(" @(" @("!C86YD:61A=&5S(#T at 6PT*(" @(" @(" @(" @
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M<R!T;R!U<V4 at 7# @87, at 82!P;&%C96AO;&1E<@T*#0H@(" @(R!+:6YD(&]F
M('5G;'D at 8G5T(&ET('=O<FMS(&9O<B!E;6)E9&1E9"!C;VUM87, at +2!$2DU#
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M#0H@(" @(" @(" @("!P87)T(#T@<W1R:6YG+G)E<&QA8V4H<&%R="P at 96]L
M+"!D96QI;6ET97(I#0H@(" @(" @(&5L<V4 at .@T*(" @(" @(" @(" @:68@
M;F]T('!A<G0 at 86YD(&D@;F]T(&EN(&QI;6ET<SH@<&%R=" ]("<B)PT*(" @
M(" @("!P87)T<UMI72 ]('!A<G0-"B @(" C(&UE<F=E(&ET(&)A8VL-"B @
M("!T>'0@/2!S=')I;F<N:F]I;BAP87)T<RP@(B(I#0H@(" @<&%R=',@/2!S
M=')I;F<N<W!L:70H='AT+"!E;VPI#0H@(" @(R!N;W<@8G)E86L at 8GD@7# -
M"B @("!F;W(@:2!I;B!R86YG92AL96XH<&%R=',I*2 Z#0H@(" @(" @(&9I
M96QD<R ]('-T<FEN9RYS<&QI="AP87)T<UMI72P@(EPP(BD-"B @(" @(" @
M<&%R='-;:5T@/2!T=7!L92AF:65L9',I#0H@(" @<F5T=7)N('!A<G1S#0H-
M"B C('5T:6QI=&EE<PT*#0ID968 at 9FEN9&QI;F5D96QI;6ET97(H='AT*2 Z
M#0H@(" @(B(B#0H@(" @<')O=FED92!S;VUE(&MB(&]F('1E>'0@=&\@=&AI
M<R!F=6YC=&EO;BX at 270@=VEL;"!D971E<FUI;F4-"B @("!T:&4 at 8F5S="!D
M96QI;6ET97(@86YD('1H97)E9F]R92!G=65S<R!T:&4@<WES=&5M#0H@(" @
M(B(B#0H@(" @;6%C(" ](")<>#!$(@T*(" @('5N:7@@/2 B7'@P02(-"B @
M("!D;W,@(#T@;6%C*W5N:7 at -"B @("!O<V5S(#T at 6V1O<RP@=6YI>"P@;6%C
M70T*(" @(", at 9FEN9"!T:&4@;VYE('=H:6-H(&=I=F5S('1H92!M;W-T(&QI
M;F5S+ at T*(" @(",@:6X at 9&]U8G0L('1H92!L;VYG97-T(&1E;&EM:71E<B!W
M:6YS+ at T*(" @(&QI<R ](%M=#0H@(" @=VAI;&4@='AT(&%N9"!T>'1;+3%=
M(&EN(&1O<SH@='AT(#T@='AT6SHM,5T@(R!#5#DW,#DP- T*(" @(&9O<B!D
M96QI;2!I;B!O<V5S. at T*(" @(" @("!L:7,N87!P96YD*"AL96XH<W1R:6YG
M+G-P;&ET*'1X="P at 9&5L:6TI*2P at 9&5L:6TI*0T*(" @(&QI<RYS;W)T*"D[
M#0H@(" @<F5T=7)N(&QI<ULM,5U;+3%=#0H-"FEF(%]?;F%M95]?/3TB7U]M
M86EN7U\B. at T*(" @(&9O<B!L(&EN("!S<&QI=%]D96QI;6ET960H<B<G)S$L
M(#(L,RP@(G9I97(B+" B<75O(B)T92(L(")E;6)E9&1E9"P at 8V]M;6$B+" B
M=&AI<PT*(" @(&ES('=I=&@@82!N97=L:6YE(BP@:&5R90T*(" @(&%N;W1H
M97(@<F5C;W)D)R<G*3H-"B @(" @<')I;G0@;&5N*&PI#0H@(" @('!R:6YT
M(&P-"@T*#0H-"FEF(%]?;F%M95]?/3TB7U]M86EN7U\B. at T*#0H@(" @9'1A
M/2<G)S$L(#(L,RP@(G9I97(B+" B<75O(B)T92(L(")T:&ES#0II<R!W:71H
M(&$@;F5W;&EN92(L(&AE<F4-"F%N;W1H97(@<F5C;W)D)R<G*C4P,# P#0H@
M(" @:6UP;W)T('1I;64-"B @("!P<FEN="!L96XH9'1A*0T*(" @('0Q/71I
M;64N=&EM92 at I#0H@(" @<W!L:71?9&5L:6UI=&5D*&1T82D-"B @("!P<FEN
8="!T:6UE+G1I;64H*2UT,0T*#0H-"@T*
`
end




From darrell at dorb.com  Thu Nov 11 18:17:30 1999
From: darrell at dorb.com (Darrell)
Date: Thu, 11 Nov 1999 18:17:30 -0500
Subject: ASCII delimited files
References: <NDBBIAPHEKGOBIHCBLIIIEOICAAA.mcfletch@vrtelecom.com>
Message-ID: <034b01bf2c9a$ed7c3830$0100a8c0@rochester.rr.com>

----- Original Message ----- 
From: Mike Fletcher <mcfletch at vrtelecom.com>
> Okay, here's an entirely different approach, extremely memory intensive,
> fairly fast. 

I'm impressed if you just whipped this up today!
Should it trim the leading empty record ???
I noticed that it returns an additional empty record.

Here's the 'C' version with your tests.

Time for file of 365 bytes
 (5 * testset, 10 records found):
 0.0
Time for file of 3650 bytes
 (50 * testset, 100 records found):
 0.0
Time for file of 36500 bytes
 (500 * testset, 1000 records found):
 0.0160000324249
Time for file of 365000 bytes
 (5000 * testset, 10000 records found):
 0.15600001812
next test may take 15 seconds or so
Time for file of 3650000 bytes
 (50000 * testset, 100000 records found):
 1.89099991322
next test may take 45 seconds or so
Time for file of 10950000 bytes
 (150000 * testset, 300000 records found):
 7.32799994946

--Darrell




From pwos at home.com  Sat Nov  6 14:40:16 1999
From: pwos at home.com (Wostenberg)
Date: Sat, 06 Nov 1999 19:40:16 GMT
Subject: Python and Smalltalk and Java envy
References: <1270663667-24391282@hypernet.com> <381ED7A6.D7D7FB13@compaq.com> <877lk1wiij.fsf@ev.netlab.co.jp> <19991102154543.A4469@optiplex.palga.uucp> <381F9EB4.63FA033F@python.net> <slrn81ve6m.l1c.neelk@brick.cswv.com> <7vqkqq$oa7$1@nnrp1.deja.com>
Message-ID: <382484F1.F8B8D712@home.com>


Doug Way wrote:

> From a historical point of view, Smalltalk is a little like Lisp, in
> that it's a little outside of the mainstream, but it has enough new
> ideas in it that will give it a very long lifespan.  Smalltalk will
> still be relevant 20 years from now, just as Lisp-based languages are
> still reasonably relevant today, after 40 years.  I suppose "having a
> lot of new ideas" and "being a little outside of the mainstream" go hand
> in hand.  (Smalltalk isn't *that* obscure, though... there are more
> Smalltalk programming jobs available than, say, Python or Lisp jobs.)

Intersting observation. Perhaps it comes down to those of us who are
comfortable with niche languages (Python, Smalltalk, Lisp) and those who
suffer Java-envy . Do Python people, Ada advocates, or little Lispers wring
there hands about the fact "we are not the most used language?" Why should
Smalltalkers?

-Alan




From garabik at melkor.dnp.fmph.uniba.sk.spam  Wed Nov  3 08:52:00 1999
From: garabik at melkor.dnp.fmph.uniba.sk.spam (Radovan Garabik)
Date: 3 Nov 1999 13:52:00 GMT
Subject: How to eliminate consol window in MS Windows
References: <3813a9be.579613@netnews.hinet.net> <3813D482.E71116AD@home.com> <VSoR3.171$Iv2.1507623@news.randori.com> <7v7897$cem$1@sunnews.cern.ch> <020601bf2112$0a9d1d60$f29b12c2@secret.pythonware.com>
Message-ID: <941640628.218028@cdwork.cvt.stuba.sk>

Fredrik Lundh <fredrik at pythonware.com> wrote:

 : you could try something like:

 : file = r"\full\path\pythonw.exe"
 : os.spawnv(os.P_NOWAIT, file, (file, "program.pyw") + tuple(args))

 : note that spawnv doesn't scan the path, and that
 : you need to pass the name of the excutable as the
 : first item in the argument tuple.

ok, thanks, it works. However...
I need to pass to program.pyw an argument, which is arbitrary string,
probably with spaces inside.
Passing is as a string passed only the first word. When I enclose
the string in quotes, it works, but when by chance there are quotes 
in orginal args, enclosing it again in quotes makes the whole
thing look ... funny.


-- 
 -----------------------------------------------------------
| Radovan Garabik http://melkor.dnp.fmph.uniba.sk/~garabik/ |
| __..--^^^--..__         garabik @ fmph.uniba.sk           |
 -----------------------------------------------------------
Antivirus alert: file .signature infected by signature virus.
Hi! I'm a signature virus! Copy me into your signature file to help me spread!



From bernhard at alpha1.csd.uwm.edu  Fri Nov 26 13:08:35 1999
From: bernhard at alpha1.csd.uwm.edu (Bernhard Reiter)
Date: 26 Nov 1999 18:08:35 GMT
Subject: XPython (X11)
References: <383E5350.1E07536A@xrce.xerox.com>
Message-ID: <slrn83tj53.k7h.bernhard@alpha1.csd.uwm.edu>

On Fri, 26 Nov 1999 10:30:56 +0100, Emmanuel Pietriga <emmanuel.pietriga at xrce.xerox.com> wrote:
>Is anyone using XPython around here?

This is the first time I heard of XPython. :)
The only reference I found is :
	http://ftp.uni-klu.ac.at/programming/python/X-extension/Top.html

>I need help, or at least links to examples (I haven't found any but the
>ones within the package)

If I want to program a GUI I look in to wxPython.
Or pyGtk or pyFOX or pyFLTK, but XPython... :)
	Bernhard
-- 
Research Assistant, Geog Dept UM-Milwaukee, USA.  (www.uwm.edu/~bernhard)
Free Software Projects and Consulting 		         (intevation.net)  
Association for a Free Informational Infrastructure            (ffii.org)



From gmcm at hypernet.com  Mon Nov 15 13:18:46 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Mon, 15 Nov 1999 13:18:46 -0500
Subject: Memory not being released?
In-Reply-To: <80pdvo$99r$1@nnrp1.deja.com>
Message-ID: <1269432569-4253767@hypernet.com>

Steve Tregidgo writes:

> I have a process (goes up to 250+ MB, cue lots of virtual memory
> usage and hence slowdown), which busies the server so much that
> it just can't bring itself to do other things -- in other words
> it's severely impractical.
> 
> The actual data that results from the process is relatively small
> (the useful stuff is less than 1MB, and other data -- such as
> objects that have been dealt with, to prevent following circular
> references -- is no more than a few MBs), and certainly well
> below the size of the actual footprint.
> 
> I've gone through the loops and recursion with a fine-tooth comb,
> and found that by eliminating certain function calls the
> footprint is reduced -- the process is thus rendered useless
> (they were the important functions, of course), but at least I
> get an idea of where the memory is going.
> 
> On the other hand, commenting out some other expressions (such as
> those that remember data -- in other words, the things that I
> would expect to take up memory and not release it again) doesn't
> make a blind bit of difference.

That's a pretty good sign that you have circular references that 
are keeping objects alive. Check out Cyclops from 
ftp://ftp.python.org/pub/python/contrib/System/.

Also, be aware that even if your python objects are getting 
collected, that doesn't mean your c runtime is giving the 
memory back to the OS. Although 250M would indicate a 
pretty horrid c runtime.
 
> So what's going on?  My guesses so far have been along the lines
> of refcounts not going down -- how else do I explain why memory
> that was allocated was seemingly not deallocated?  I've tried
> del'ing everything in sight, storing ids instead of objects,
> inserting the odd sys.exc_traceback=None to clear those pesky
> traceback reference holders ... none of it has helped so far.
> 
> Are there other hidden things that might hold references to my
> objects?  And if so, what can I do about them?  After using an
> object once, I don't want (don't need) to keep it in memory any
> more, so for those objects I'll be revisiting at a later date, I
> just remember their id and delete them.

In your code, dict1 is holding a reference to every object 
you've visited, and thus keeping them all alive. OTOH, judging 
by "for item in obj.subs:", it looks like the whole graph is alive 
all the time anyway. Essentially, you've got a huge pointer-
based data structure. You need to hoist most of the smarts up 
to a Graph object, and use some kind of persistent key to 
identify objects (which are really on disk, or recreated on the 
fly...). So that snippet might be:
  for itemname in obj.subs:
    obj = graph.getobject(itemname)
 
> If it's any help, the process does something like this:
> 
> def recurse(obj, list, dict1, dict2):
>   # Prevent circular recursion...
>   if dict1.has_key(obj.__id__):
>     return
>   else:
>     dict1[obj.__id__] = None
> 
>   a = do_thing(obj, dict2)
> 
>   # Recurse into obj's children, or do something else
>   for item in obj.subs:
>     if item.spam:
>       recurse(item, list, dict1, dict2)
>     else:
>       b = do_thing(item, dict2)
> 
>   c = do_thing(obj, dict2)
> 
>   # Remember some things -- commenting
>   # out produces no memory saving.
>   list.append(a,b,c)
> 
> The function do_thing implements a loop; in total (over the whole
> process) the loop's body is executed tens of thousands of times
> -- it causes the biggest memory saving when commented out (and of
> course it's the most important bit), but does no obvious
> "remembering" of things.
> 
> Does anybody have any ideas?  Or can you maybe point me to some
> documentation that deals with this -- the FAQ looked promising
> but ultimately didn't help.
> 
> Cheers,
> Steve
> 
> --
>        -- Steve Tregidgo --
> Developer for Business Collaborator
>         www.enviros.com/bc
> 
> 
> Sent via Deja.com http://www.deja.com/
> Before you buy.
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list



- Gordon



From fredrik at pythonware.com  Fri Nov 19 05:56:08 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 19 Nov 1999 11:56:08 +0100
Subject: Batch window
References: <383525CB.6E2EE3AF@obop.com.pl>
Message-ID: <016c01bf327c$b2ad5260$f29b12c2@secret.pythonware.com>

Przemyslaw G. Gawronski <gawron at obop.com.pl> wrote:
> Is it posible to automaticaly minimize the python window, when starting
> a GUI application on Windows ?

you can get rid of it entirely, by using
the "pythonw" interpreter:

    pythonw myapp.py

(note that stdout and strerr are directed
to the bit bucket)

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
(news: just posted an errata for the 'first printing') -->




From LCortes at Flash.Net  Sat Nov 27 13:52:14 1999
From: LCortes at Flash.Net (Luis Cortes)
Date: Sat, 27 Nov 1999 18:52:14 GMT
Subject: Prolog Extension
Message-ID: <19991127.19045200@negrita.brunner.org>

Is there a version of Python with Prolog Extension??
Where can I find it if there is?

Thanks,
Luis.

P.S. Could you please reply to the sender too.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/19991127/7ddb2e60/attachment.html>

From hnowak at cuci.nl  Fri Nov 19 17:20:42 1999
From: hnowak at cuci.nl (Hans Nowak)
Date: Fri, 19 Nov 1999 23:20:42 +0100
Subject: Python in DOS
In-Reply-To: <DB6390A9770ED211B2440000F807FC7B01039D69@nt_exlxe.lxe.com>
Message-ID: <199911192222.XAA26465@dionysus.fw.cuci.nl>

On 19 Nov 99, at 10:59, Curtis, Craig M. wrote:

> I tried setting PYTHONCASEOK to 0 and did the configure and make and still
> have the same problem.  Any more suggestions?

Yeah, try setting it to 1. :o) 0 doesn't have much effect AFAIK. This 
variable does not affect compilation though, it only affects the 
interpreter when it runs.

If that doesn't help, do as I said... uncomment the code. I think 
uncommenting a coupla fprintf statements in getpath.c will do. You'll 
know 'em when you see 'em... they're these infamous error/warning 
messages of course.

HTH,

--Hans Nowak (zephyrfalcon at hvision.nl)
Homepage: http://fly.to/zephyrfalcon
You call me a masterless man. You are wrong. I am my own master.



From pinard at iro.umontreal.ca  Mon Nov  1 19:56:49 1999
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 01 Nov 1999 19:56:49 -0500
Subject: Default indentation of 4 columns
In-Reply-To: "Mark Hammond"'s message of "Mon, 25 Oct 1999 16:17:22 +1000"
References: <004001bf1eb0$99d713f0$0501a8c0@bobcat>
Message-ID: <oqaeoxhfse.fsf@titan.progiciels-bpi.ca>

"Mark Hammond" <mhammond at skippinet.com.au> writes:

> However, look at "string.split.__doc__" - this does use a style
> similar to what Tim described and it is in .py code - and indeed does
> end up showing the arg information twice in the calltip window.

> So-basically-I-have-no-idea-what-Tim-meant<wink>-ly,

Hmph!  I see:


Python 1.5.2 (#1, Jul 23 1999, 06:38:16)  [GCC egcs-2.91.66 19990314/Linux (egcs- on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> print string.split.__doc__
split(str [,sep [,maxsplit]]) -> list of strings
splitfields(str [,sep [,maxsplit]]) -> list of strings

Return a list of the words in the string s, using sep as the
delimiter string.  If maxsplit is nonzero, splits into at most
maxsplit words If sep is not specified, any whitespace string
is a separator.  Maxsplit defaults to 0.

(split and splitfields are synonymous)


You see that `str' is used in the prototype, while `s' is used in the
description.  Such errors might be caught more easily by the eye if the
capitals convention was followed.  Compare (I'm using simili-GNU style):


split(STR [,SEP [,MAXSPLIT]]) -> list of strings
splitfields(STR [,SEP [,MAXSPLIT]]) -> list of strings

Return a list of the words in the string STR, using SEP as the
delimiter string.  If MAXSPLIT is nonzero, splits into at most
MAXSPLIT words.  If SEP is not specified, any whitespace string
is a separator.  MAXSPLIT defaults to 0.

(`split' and `splitfields' are synonymous.)


P.S. - If I were doing it, I would also write STRING and SEPARATOR in full.
Using smallish abbreviations come from prehistory :-) :-).

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





From ivnowa at hvision.nl  Mon Nov 22 16:23:07 1999
From: ivnowa at hvision.nl (Hans Nowak)
Date: Mon, 22 Nov 1999 22:23:07 +0100
Subject: Python in DOS / PYTHONHOME
In-Reply-To: <DB6390A9770ED211B2440000F807FC7B0103A127@nt_exlxe.lxe.com>
Message-ID: <199911222122.WAA22314@axil.hvision.nl>

On 22 Nov 99, Curtis, Craig M. wrote:

> Commenting out those dreaded printfs certainly does work :-)
> 
> I guess I do not understand what PYTHONHOME should be set to.  I tried
> setting it to the python-1.5.2 root and also to the lib under it, but
> neither of these work.

I don't understand PYTHONHOME either. ;-)  That is, I never tried to 
use it.

Running 'python -?' yields:

[...]
PYTHONHOME   : alternate <prefix> directory (or 
<prefix>;<exec_prefix>). The default module search path uses 
<prefix>/python1.5.

Maybe this makes more sense in a Unix environment? I can imagine Unix 
directories called usr/lib/python1.5 or something, but in DOS, this 
is quite uncommon. 

Does anyone use this under DOS?

--Hans Nowak (zephyrfalcon at hvision.nl)
Homepage: http://fly.to/zephyrfalcon
You call me a masterless man. You are wrong. I am my own master.



From claudius at catlover.com  Fri Nov  5 23:15:59 1999
From: claudius at catlover.com (claudius at catlover.com)
Date: Sat, 06 Nov 1999 04:15:59 GMT
Subject: Numbers in SQL statements
References: <38235836.D2F9A909@src.uchicago.edu>
Message-ID: <3YNU3.17551$Ec1.184623@news1.frmt1.sfba.home.com>

Nick Collier <nick at src.uchicago.edu> says:
>Hi,
>
>I've been creating SQL statements from dictionaries holding the field
>name and the new value to insert or update. As might be expected this
>new value is often a number. When turning this number into a string in
>order to create the sql statement the number will have a "L" appended if
>it is a long. Consequently, the resulting SQL might look like the
>following:
>
>update myTable
>set id_num = 23L
>where food = 'spam'
>
>I've been testing if the value is a number whose string representation
>ends with "L" and then stripping the "L", but am pretty sure there must
>be a better way. If so, any suggestions as to what is?

Try (if you're using a module that supports the full Python DB API 1.0):

cursor.execute('update myTable set id_num = ? where food = ?', (id_num, food))

This mechanism is much more forgiving than stringifying directly into the
query.

I've been using this format with great success (including storing binary
pickled datastructures and marshalled Python bytecode in the database.  Whee!)



From jima at aspectdv.com  Tue Nov 16 15:28:24 1999
From: jima at aspectdv.com (Jim Althoff)
Date: Tue, 16 Nov 1999 12:28:24 -0800
Subject: Beginning programming with Python
In-Reply-To: <38317787.19AFDAF5@compaq.com>
References: <sng0ychgwt.fsf@motorola.com>
 <382B3A53.5816B2@callware.com>
 <382B3D79.74CD3E0F@email.sps.mot.com>
 <80p9ds$ppo$1@newshost.accu.uu.nl>
 <asteroidal-80psfo/INN-2.2.1/bat@broadway.news.is-europe.net>
Message-ID: <4.2.0.58.19991116121641.00b4bd60@mail.aspectdv.com>

yes. yes. yes.

(If only we didn't have to type "self" into
every method arg list. sigh.)

Jim


At 04:25 AM 11/17/99 +1300, Greg Ewing wrote:
>Janos Blazi wrote:
> >
> > the indenting this time? There is actually nothing else you can forget in
> > Python.
>
>Well, not quite - you can forget the "self" in method
>declarations fairly easily...
>
>Greg
>
>--
>http://www.python.org/mailman/listinfo/python-list





From abrahams at mediaone.net  Sun Nov  7 17:52:48 1999
From: abrahams at mediaone.net (Dave Abrahams)
Date: Sun, 07 Nov 1999 17:52:48 -0500
Subject: Python and generating images for web
References: <3824F785.40DF4393@yahoo.com>
Message-ID: <xhnV3.3468$x12.141830@ndnws01.ne.mediaone.net>

In article <3824F785.40DF4393 at yahoo.com> , Alexander Stepanov 
<a_stepanov at yahoo.com>  wrote:

> (I am in a process of comparison Perl and Python to solve this problem.
> I plan to use the application
> under WindowsNT4.0 environment + Perl? or Python interp-r).

Hi Alexander!

Unfortunately, I can't help you with your gif/png/chart/plot problems,
though I'm sure someone will step forward with a clue, since so many useful
libraries have been written in Python. I assume you've tried the search
engine at www.python.org already. Also I have nothing to say about Perl,
having never used it.

I just wanted to express my appreciation of coding in Python for its
parallels with generic programming in C++, something I have a hunch you may
take a liking to also. While you don't get overloading ,or the performance
advantages, the similarities in code structure keep jumping out at me. In
particular the lack of strong type checking makes _everything_ generic. The
ability to pass type objects as function parameters is analogous to type
template parameters.

I have to say that once I got over my initial revulsion at the lack of
compile-time type checking, I've quite enjoyed my experience with Python. I
hope you have as much fun as I have.

Best Regards,
Dave Abrahams



From volumeregeling at hotmail.com  Mon Nov 29 06:22:18 1999
From: volumeregeling at hotmail.com (volumeregeling)
Date: Mon, 29 Nov 1999 11:22:18 GMT
Subject: help with opening html files
Message-ID: <384261c4.8540090@122.0.0.250>

Hi,

Can anyone tell me how to open a html file from within python in
windows.

thanks



From timmuddletin at news.vex.net  Thu Nov 25 00:39:39 1999
From: timmuddletin at news.vex.net (tim muddletin)
Date: 25 Nov 1999 05:39:39 GMT
Subject: How to perform syntax check in Python
References: <382A2384.CC37FBA8@hp.com> <00bf01bf2c1d$2a927be0$f29b12c2@secret.pythonware.com> <3838CE1A.C7BE0530@horvath.com> <slrn83hl31.p19.wtanksle@hawking.armored.net> <38398D45.DFA52EFD@horvath.com> <383AD4F1.740DFD7E@computer.org>
Message-ID: <81ii2r$2pq9$1@hub.org>

On Tue, 23 Nov 1999 09:54:57 -0800, Charles Y. Choi wrote:
>Look for trace.py by Andrew Dalke.

I tracked this down to 
ftp://starship.python.net/pub/crew/dalke

So i could enter it into my Vaults of Parnassus database
( http://localhost/~x/parnassus/apyllo.py?i=63889258 )

My problem is that i don't really understand what this script does.
There is no general description provided. I suppose it's one of those
things that is self-evident to those who know what it is. <-:

What is this "coverage" stuff? Can someone provide me with a description 
for the database? I'm lost on this one; my ignorance is showing.

(ps. also it is announced in usenet --- also without a general 
description -- here: http://x33.deja.com/=dnc/getdoc.xp?AN=49968748 )







From Gareth.McCaughan at pobox.com  Fri Nov 26 17:36:36 1999
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: 26 Nov 1999 22:36:36 +0000
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <867lj85g4c.fsf@g.local>
	<383CC835.B585EBDD@Lugoj.Com> <m3ogcjrn41.fsf@atrus.jesus.cam.ac.uk>
	<383D9C63.3A7E5806@Lugoj.Com> <86ogci2tfm.fsf@g.local>
	<al8g0xt5k8f.fsf@sirppi.helsinki.fi>
Message-ID: <86aeo0283v.fsf@g.local>

Markus Stenberg wrote:

[I said:]
>> Garbage collectors have got a *lot* better since the days when
>> GC meant half-second pauses waiting for the machine to sweep
>> through all its memory. Let alone 20 seconds.
> 
> Where can I admire such a beast? At least none of the Java
> implementations(*) I have tried(**) have one. say, for 128MB of trash, it's
> consistent N second sweep (0.5-10, depending on hardware).

In a modern Lisp implementation. Modern Smalltalks are probably
just as good. (Unfortunately, "modern" might have to imply
"commercial" here; I'm not sure of the quality of the GC in
the various free implementations of those languages.)

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
sig under construction



From fredrik at pythonware.com  Thu Nov  4 02:54:40 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 4 Nov 1999 08:54:40 +0100
Subject: Python and Literate Programming
References: <Pine.LNX.3.96.991104011425.943A-100000@dhcp9545112>
Message-ID: <006501bf2699$da9cbca0$f29b12c2@secret.pythonware.com>

Philip H. Jensen <pjensen at columbus.rr.com> wrote:
> Incidentally, I'm also a booster for Modula-3, and of course Python
> has well-known and intentional similarities to M3.  But one of my other
> enthusiasms is Literate Programming, and I'm wondering about the
> possibilities for it in Python.

you should be able to find something called
interscript here:

    http://www.triode.net.au/~skaller

(that link doesn't work when I'm writing this;
if it has moved, you might be able to locate
it via:

    http://www.ozemail.com.au/~skaller/

</F>

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




From NebtShpilfoygel at news.vex.net  Sun Nov 14 00:32:05 1999
From: NebtShpilfoygel at news.vex.net (Nebt Shpilfoygel)
Date: 14 Nov 1999 05:32:05 GMT
Subject: NT admin scripts with python
References: <000201bf2e34$26d93750$020a10ac@p350.gnaa.net>
Message-ID: <80lhgl$22gn$1@hub.org>

On Sat, 13 Nov 1999 16:06:50 -0800, geoff nordli <geoff at gnaa.net> wrote:
>Does anyone know if there is a reservoir of python scripts on the net.  Kind
>of like Perl's CPAN

It's not like CPAN but you might try http://www.vex.net/~x/parnassus/

...




From mwh21 at cam.ac.uk  Tue Nov 16 12:04:43 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 16 Nov 1999 17:04:43 +0000
Subject: __str__ vs. __repr__
References: <000001bf275c$02370d20$1d2d153f@tim> <01fd01bf2dce$d2c03cf0$f29b12c2@secret.pythonware.com> <942499879snz@vision25.demon.co.uk> <m3n1shp2uu.fsf@atrus.jesus.cam.ac.uk> <B65809622FDC7452.143B25D04FC64B07.A4A6F24F2451314F@lp.airnews.net> <slrn83136e.tv9.wtanksle@hawking.armored.net>
Message-ID: <m3vh721i6s.fsf@atrus.jesus.cam.ac.uk>

wtanksle at hawking.armored.net (William Tanksley) writes:

> On 15 Nov 1999 20:48:39 GMT, Tres Seaver wrote:
> >Michael Hudson  <mwh21 at cam.ac.uk> wrote:
> >>philh at vision25.demon.co.uk (Phil Hunt) writes:
> 
> >>It's one of those things I might get round to writing one of these
> >>days. I don't really know how hard it would be. Not especially, I
> >>suspect.
> 
> >Besides, you need to defer it until _after_ this year's Pythonic award,
> >for which you are already in pole position for inventing bytecodehacks :).
> 
> You might note that Tim is rather proud of being the only living recipient
> of the Pythonic award (for some definition of 'life').  I understand he
> might take action to keep that distinction if someone else were to win the
> award.
> 
> If you get what I mean.
> 

I don't think the bytecodehacks, while sufficiently dark and useless
to be a tim-ism, qualify me in any way for a Pythonic Wizard Hat when
compared the sheer contribution in terms of wit, code, accurate
advice, experience, emacs modes and general amazingly useful stuff
that Tim Peters has given to the Python community.

<0.0 wink>.

Try http://www.python.org/tim_one/ if you don't see what I mean.

not-a-bot-yet-ly y'rs - Michael



From guido at cnri.reston.va.us  Tue Nov  2 20:51:13 1999
From: guido at cnri.reston.va.us (Guido van Rossum)
Date: 02 Nov 1999 20:51:13 -0500
Subject: Can't start IDLE
References: <381F4EB1.AFD9BA32@src.uchicago.edu>
Message-ID: <5lg0yotka6.fsf@eric.cnri.reston.va.us>

Nick Collier <nick at src.uchicago.edu> writes:

> On winNT sp3 I get the following message when I do a "python idle.py"
> from the command line:
> 
> C:\Program Files\Python\Tools\idle>python idle.py
> Traceback (innermost last):
>   File "idle.py", line 3, in ?
>     PyShell.main()
>   File "PyShell.py", line 612, in main
>     fixwordbreaks(root)
>   File "EditorWindow.py", line 606, in fixwordbreaks
>     tk.call('tcl_wordBreakAfter', 'a b', 0) # make sure word.tcl is
> loaded
> TclError: invalid command name "tcl_wordBreakAfter"
> 
> I looked around deja.com and found one other person with the same
> problem, but no response so I'm asking here.

Could it be that you have an older (or newer) Tcl version installed?
Python (for now, until I upgrade the _tkinter module) requires Tcl/Tk
8.0.5 -- which is the version that comes with the Python 1.5.2 for
Windows installer.

Try to erase all traces of other Tcl installs and reinstall Python and
see if that works.

--Guido van Rossum (home page: http://www.python.org/~guido/)



From g_will at cyberus.ca  Mon Nov 29 12:08:40 1999
From: g_will at cyberus.ca (Gordon Williams)
Date: Mon, 29 Nov 1999 17:08:40 GMT
Subject: Pythonwin Reset
References: <01bf38df$8127e0c0$3a42c3d1@Cyberus.cyberus.ca> <_X004.7432$VA6.38559@news-server.bigpond.net.au> <01bf39bd$a28831a0$4155c3d1@Cyberus.cyberus.ca> <81rmh5$9e$1@hub.org>
Message-ID: <01bf3a8c$3f855b80$2241c3d1@Cyberus.cyberus.ca>

Tim,

Thanks for the note.

It now works in that it eliminates all the names, but it doesnt quite fix
my problem.  When I make a GUI with wxPython (using Pythonwin IDE) and I
get an exception, the next time I try and run the program after fixing the
mistake, I get another exception saying One wxApps process only!  Something
was left behind.  The only way I have found to clear the problem was to
close Pythonwin and then restart it and reload the file.

Regards,

Gordon Williams



tim muddletin <tim muddletin at news.vex.net> wrote in article
<81rmh5$9e$1 at hub.org>...
> On Sun, 28 Nov 1999 16:29:35 GMT, Gordon Williams <g_will at cyberus.ca>
wrote:
> >NameError: __main__
> >
> >what has gone wrong?
> 
> A non-obvious gotcha. You have to "import __main__" first, to get access
> to the namespace in that way.
> 
> 
> 



From moonseeker at my-deja.com  Thu Nov 25 08:57:23 1999
From: moonseeker at my-deja.com (moonseeker at my-deja.com)
Date: Thu, 25 Nov 1999 13:57:23 GMT
Subject: natural speech parser?
Message-ID: <81jf82$kil$1@nnrp1.deja.com>

Hi,

I'm looking for a command parser to parse sentences like:

enable KEY1 and write 'Hello World' to var2

I havn't done such a thing like parsers or compilers (which works with
command trees) since yet and would welcome any suggestion.

Mike


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



From T.Middleton at news.vex.net  Fri Nov 19 04:43:25 1999
From: T.Middleton at news.vex.net (T. Middleton)
Date: 19 Nov 1999 09:43:25 GMT
Subject: Proposal: Python Info Collective (another ramble)
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <JlFX3.3659$s87.67930@dfiatx1-snr1.gtei.net> <slrn82unnr.mja.thor@localhost.localdomain> <80ubrg$tuj$1@hub.org> <3dso23jvmf.fsf@amarok.cnri.reston.va.us>
Message-ID: <81363t$1u17$1@hub.org>

On 18 Nov 1999 17:14:00 -0500, Andrew M. Kuchling wrote:
>it to a host without a weekly bandwidth limitation.  Someone mentioned
>putting it onto starship.python.net; is anything happening with that?

I will be confirming by this weekend (hopefully) whether Vaults of Parnassus
stays on vex.net (with bandwidth expanded) or not. Surely by the end of
the month at the most all these things will be sorted out.

(For the curious. Been doing about 30 megs traffic a day since it was 
'outed' a mere 4 or 5 days ago... have recieved 11 hundred hits on the 
main page.)

...




From mal at lemburg.com  Fri Nov  5 08:31:47 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Fri, 05 Nov 1999 14:31:47 +0100
Subject: Module for generic platform information: platform.py
References: <199911042207.XAA16602@axil.hvision.nl>
Message-ID: <3822DC43.4B022439@lemburg.com>

Hans Nowak wrote:
> 
> On 4 Nov 99, M.-A. Lemburg wrote:
> 
> > There is a new release of the platform.py module available at:
> >
> >         http://starship.skyport.net/~lemburg/platform.py
> >
> > Still needed:
> >     ? support for WinCE
> >     ? support for MS-DOS (PythonDX ?)
> 
> Hmm... Doesn't that work already? I tried it several days ago, so
> with an older version of platform.py, obviously. Here's what it says
> when fed to Python-DX:
> 
> c:/djgpp/bin/uname.exe: invalid option -- p
> Try `c:/djgpp/bin/uname.exe --help' for more information.
> MS-DOS-7-pc
> 
> (OK, the error message doesn't belong there, but that's just because
> of the uname call. Most DOS users won't have an uname at all on their
> system. In spite of this, it does find the right DOS version. By the
> way, note that the os.popen, used to call uname, works. ;-)

Not bad :-) I wonder where the 'pc' in the string comes from...

BTW, I've added an if to prevent the call to 'uname -p' on DOS
platforms, so the above error should no longer show up.
 
> When run with Python for Windows (on the same PC) it says:
> 
> Windows-95-4.00.1111
> 
> I don't know what it does on DOS 6 and lower, though. =/

Does Python run on those ?

-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                    56 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/





From aa8vb at yahoo.com  Wed Nov  3 07:17:27 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Wed, 3 Nov 1999 07:17:27 -0500
Subject: Exceptions from callbacks
In-Reply-To: <381FAE6D.73F426FE@lords.com>
References: <381CA8E5.5D69E64@lords.com> <7virj4$b7b$1@towncrier.cc.monash.edu.au> <381FAE6D.73F426FE@lords.com>
Message-ID: <19991103071727.A87523@vislab.epa.gov>

Eric Dorland:
 |Jonathan Giddy wrote:
 |> Eric Dorland <dorland at lords.com> writes:
 |> 
 |> >I'm writing a Python wrapper for a library. The library stores some
 |> >callbacks for some events, and with some glue code, get these to call
 |> >python functions. That's the easy part. Now since its the library these
 |> >callback and not Python per se, how do you pass an exception back to the
 |> >interpreter, because you can't return NULL from the callback. Do you
 |> >just set the exception, and Python will pick up on it?
...
 |You're not understanding exactly. It is synchronous (ie one thread), but
 |since the library is calling the callback, and not the python
 |interpreter, I can't return NULL to the interpreter. Basically, is there
 |a way from inside the callback to say "Wake up python, an exception!",
 |without having to return NULL to the interpreter, which is impossible
 |from the callback.

I've done this, and I got the impression that Jonathan understood your
question.  What he described works and it's what I'm using.

To extend your description, you don't say "Wake up, Python!" from the
callback.  You say it from the C function(s) from which you entered the C
library that invoked the callback.  Return NULL, and the exception is
restarted with the Python stack frames saved in the callback in-tow.

To make this concrete, for a C library API that might potentially invoke a
callback (which in-turn would relay to Python) would look like this:

  (Python) -> HandlePendingMessages -> MYLIB -> Callback -> (Python)

And your wrapper code might look like this:

------------------------------------------------------------------------------
  PyObject *_wrap_MYLIBHandlePendingMessages( PyObject *self, PyObject *args )
  {
    PyErr_Clear();

    <<< call C function here >>>

    /*  If a Python callback threw an exception, restart its propagation  */
    /*    in the controlling Python.                                      */
    if ( PyErr_Occurred() )
      return NULL;
    
    ...
  }

------------------------------------------------------------------------------
  static void _wrap_MYLIBCallback( void )
    /*  C callback used to relay to Python callable objects.  */
  {
    ...

    /*  Pass the buck onto Python  */
    pyresult = PyEval_CallObject( pyfunc, pyargs );     /*  Call Python     */
    Py_DECREF( pyargs );                                /*  Trash arg list  */
    if ( pyresult )
      Py_DECREF( pyresult );                            /*  Trash any result*/

    /*  If the callback threw an exception, it will be propagated across    */
    /*    the C layer ( we check PyErr_Occurred() on the other side).       */
  }
------------------------------------------------------------------------------

If the Python callback tossed an exception, then it's still lying around
when _wrap_MYLIBCallback() returns.  We then return through the C library,
which eventually returns to _wrap_MYLIBHandlePendingMessages, who
recognizes that a Python exception is pending and restarts its propagation
in the controlling Python stack frames.

Note that if you have a C library layer that isn't Python-exception-aware,
then it could go on to perform useless or invalid work after a Python
callback throws an exception.  In this case you'll need to "make" it aware.
That is, have it call PyErr_Occurred(), or turn Python exceptions into C
library errors inside of your callback wrappers.

Hope this clarifies things a bit.

Randall



From JamesL at Lugoj.Com  Thu Nov 25 14:48:09 1999
From: JamesL at Lugoj.Com (James Logajan)
Date: Thu, 25 Nov 1999 11:48:09 -0800
Subject: Python complaints
References: <Pine.LNX.4.21.9911251159140.11528-100000@fep132.fep.ru>
Message-ID: <383D9279.98543836@Lugoj.Com>

Oleg Broytmann wrote:
> 
> On Wed, 24 Nov 1999, James Logajan wrote:
> > def ifexp(cond, trueVal, falseVal):
> >    if cond:
> >       return trueVal
> >    else:
> >       return falseVal
> >
> > print "You scored %s point%s" % (score, ifexp(score == 1, "", "s"))
> 
>    The problem with this solution is that python will calculate both
> expressions (trueVal, falseVal) even if "cond" is false - what is VERY
> undesireable in most situations. What if I just cannot calculate falseVal
> if "cond" is false?
> 

Then how about:

if score == 1: suffix = ""
else: suffix = "s"
print "You scored %s point%s" % (score, suffix)

I better put a <wink> in here; never know when the terminally lazy might
become terminally ballistic! Besides the pluralization algorithm isn't
portable to other languages (such as German...(do I need to add another
wink?))



From wtanksle at hawking.armored.net  Mon Nov 15 17:43:38 1999
From: wtanksle at hawking.armored.net (William Tanksley)
Date: 15 Nov 1999 22:43:38 GMT
Subject: __str__ vs. __repr__
References: <000001bf275c$02370d20$1d2d153f@tim> <01fd01bf2dce$d2c03cf0$f29b12c2@secret.pythonware.com> <942499879snz@vision25.demon.co.uk> <m3n1shp2uu.fsf@atrus.jesus.cam.ac.uk> <B65809622FDC7452.143B25D04FC64B07.A4A6F24F2451314F@lp.airnews.net>
Message-ID: <slrn83136e.tv9.wtanksle@hawking.armored.net>

On 15 Nov 1999 20:48:39 GMT, Tres Seaver wrote:
>Michael Hudson  <mwh21 at cam.ac.uk> wrote:
>>philh at vision25.demon.co.uk (Phil Hunt) writes:

>>It's one of those things I might get round to writing one of these
>>days. I don't really know how hard it would be. Not especially, I
>>suspect.

>Besides, you need to defer it until _after_ this year's Pythonic award,
>for which you are already in pole position for inventing bytecodehacks :).

You might note that Tim is rather proud of being the only living recipient
of the Pythonic award (for some definition of 'life').  I understand he
might take action to keep that distinction if someone else were to win the
award.

If you get what I mean.

>Tres.

watching-my-back-and-underachieving-ly y'rs,
-- 
-William "Billy" Tanksley
  <1.0 wink>



From jeremy at cnri.reston.va.us  Mon Nov  1 17:04:03 1999
From: jeremy at cnri.reston.va.us (Jeremy Hylton)
Date: Mon, 1 Nov 1999 17:04:03 -0500 (EST)
Subject: Feature request for interactive interpreter
In-Reply-To: <51C9F3C392D0D111A15600A0C99E64A9AF5606@localhost>
References: <51C9F3C392D0D111A15600A0C99E64A9AF5606@localhost>
Message-ID: <14366.3667.934145.165288@goon.cnri.reston.va.us>

Python already lets you run user-supplied code every time you fire up
an interactive interpreter (PYTHONSTARTUP environment variable).  If
you add your definition for help that way, there would be no need to
change Python.


Jeremy




From donn at u.washington.edu  Wed Nov 10 13:27:14 1999
From: donn at u.washington.edu (Donn Cave)
Date: 10 Nov 1999 18:27:14 GMT
Subject: Exceptions: ^z
References: <38299522.5D06E8D5@sw.ods.com>
Message-ID: <80cde2$oou$1@nntp6.u.washington.edu>

Quoth Sai Ramanath <sai at sw.ods.com>:
| Is there a way to catch Ctrl-Z which in UNIX lets the user suspend the Python
| shell into background?
|
| I know we can catch Ctrl-C and Ctrl-D using KeyboardInterrupt and EOFError
| respectively.

You'd use signal.signal(signal.SIGTSTP, signal_handler_function)

I wouldn't want to actually do that, myself.  Partly because UNIX signal
handling and Python don't get along too well, partly because it doesn't
pay to tinker with the delicate machinery of Berkeley job control.

If all you really want is to turn off this feature, do that with
the terminal driver:

  import termios
  import TERMIOS

  attr = termios.getattr(0)
  attr[6][TERMIOS.VSUSP] = '\000'   # '\032' sets it back to ^Z
  termios.setattr(0, TERMIOS.TCSANOW, attr)

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu



From a.eyre at optichrome.com  Mon Nov 29 09:15:45 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Mon, 29 Nov 1999 14:15:45 -0000
Subject: XML tool status?
In-Reply-To: <DIv04.8846$cz3.320849@nnrp1.uunet.ca>
Message-ID: <000d01bf3a74$3a4aa110$3acbd9c2@peridot.optichrome.com>

> 1) I can't find a Windows version of the software; and

Odd. It should be platform independent.

> 2) the Python XML SIG status page has a "most recent update" date from
> December 1998 --- is that fossilware, or has there really not 
> been anything new posted since then?

The website is not updated as often as the source. You'll have to grab it
via CVS. See: http://www.python.org/sigs/xml-sig/anon-cvs.html

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From gmcm at hypernet.com  Mon Nov  1 15:40:01 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Mon, 1 Nov 1999 15:40:01 -0500
Subject: __str__ vs. __repr__
In-Reply-To: <19991101121709.A134039715@vislab.epa.gov>
Message-ID: <1270634443-26149069@hypernet.com>

Randall Hopper wrote:

> 1) Is there a convention for what __str__ and __repr__ should
> return for
>    classes?
> 2) Or, whatever they return, should they return the same value?
> 3) If so, why have both in the language?

If possible, __repr__ should return something that when eval'd 
yields an identical object. If that's not possible, it should be 
the programmer's view of the object.

__str__ is free to return a highly munged user-friendly string. 
str(obj) will use obj.__repr__ if obj.__str__ doesn't exist.

Consider this:

class HTMLTable:
   blah blah blah

t = HTMLTable(...)
repr(t) # -> "MyModule.HTMLTable([...])"
str(t) # -> "<TABLE><TR>....."

(see IPWP for this idea carried to extreme lengths).

>      Searching the archives yields an example with odd behavior
>      that
> suggests they should generally return the same value:
> 
>    class c:
>        def __str__(self):  return 'foo'
>        def __repr__(self): return 'bar'
> 
>    >>> str(c())
>    'foo'
> 
>    >>> str([c(), c(), c()])
>    '[bar, bar, bar]'

Sick enough, but not convoluted enough, to be a Tim-ism...

- Gordon



From T.Muddletyn at news.vex.net  Tue Nov 16 01:58:14 1999
From: T.Muddletyn at news.vex.net (T. Muddletyn)
Date: 16 Nov 1999 06:58:14 GMT
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <80pe77$osc$1@hub.org> <slrn830he9.n28.mgm@unpkhswm04.bscc.bls.com>
Message-ID: <80qva6$1fgh$1@hub.org>

On 15 Nov 1999 17:41:30 GMT, Mitchell Morris wrote:
>My personal take is that you never actually look at CPAN in a browser.  The

Yes, you make a good point there. My rant was based on a narrow and 
biggotted browser view. 

The command line tools for searching and retrieving are a potentially very
handy thing. Martijn Faassen in a previous message made the great suggestion of
possibly integrating Parnassus somehow with the "distutils" system which is
being worked on. It would be very interesting. 

As things are now, with python resources being so diversely packaged and
hosted, it would require a fair bit of effort and time by all owners of
resources to re-package themselves for compliance for a full blown system like
this to take off (though one has to start somewhere!)

On the other hand, a simple command-line querying util, at every least, would
be really cool (and save me a bit of bandwidth too! heh) I'll definitely
think about this!

...




From da at ski.org  Fri Nov  5 13:40:36 1999
From: da at ski.org (David Ascher)
Date: Fri, 5 Nov 1999 10:40:36 -0800 (Pacific Standard Time)
Subject: IPC8: SouthWest Airlines Internet specials
Message-ID: <Pine.WNT.4.04.9911051036270.260-100000@rigoletto.ski.org>

For those of you who are paying for the IPC8 conference out of your
personal entertainment budget: I just bought tickets to go to the IPC8
Conference for $194 roundtrip on Southwest from Oakland, CA.  I fly into
Baltimore-Washington, which saves a *lot* of money, even factoring in
ground transportation. I mention it because the internet special has to be
bought before November 8th.

--david ascher 






From matz at netlab.co.jp  Thu Nov  4 23:30:43 1999
From: matz at netlab.co.jp (Yukihiro Matsumoto)
Date: 05 Nov 1999 13:30:43 +0900
Subject: Some more Ruby observations (Re: Comparison between Python and "Ruby")
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp>
 <7vjtrc$hmm$1@newshost.accu.uu.nl> <3821A860.5D32AFCC@compaq.com>
 <slrn824eo2.lsi.neelk@brick.cswv.com>
Message-ID: <87904dv9u4.fsf@ev.netlab.co.jp>

neelk at brick.cswv.com (Neel Krishnaswami) writes:

|Are there any syntax-tree munging phases to do stuff like common
|subexpression elimination and loop hoisting? I guess that could yield
|enough speedups to make a VM not yet necessary.

No.  It's pretty hard to optimize a dynamic language like Ruby, in
which even integer addition can be re-defined.  The optimizer can't
tell whether a subexpression has side effect or not.

Optimization ideas are welcome, though. :-)
							matz.



From news at dorb.com  Mon Nov 15 14:19:01 1999
From: news at dorb.com (Darrell)
Date: Mon, 15 Nov 1999 14:19:01 -0500
Subject: Memory not being released?
References: <80pdvo$99r$1@nnrp1.deja.com>
Message-ID: <%_YX3.8353$wG3.451084@typhoon.nyroc.rr.com>

Here's an idiom I like in this case.
Store your objects in a dict based on id(obj), just as you've done.
Now if objects must reference each other use the id(obj)
and look up the obj. And delete them from the dict when needed.
Yes it's slow but only use it in these big ugly cyclic data structures.

But notice how fast memory gets used up.
This eats 84meg.
>>> dict={}
>>> for x in xrange(1000000):
...     dict[x]=(x,x+1)
...
>>>

You might take a look at shelve. It gives you a persistent dictionary.
I've been playing with a flat file manager that can reclaim freed space.

--
--Darrell
Steve Tregidgo <smst at bigfoot.com> wrote in message
news:80pdvo$99r$1 at nnrp1.deja.com...
> Hi there,
>
> I'm having problems with a large process, and I'd really appreciate
> some help in tracking down the cause of its bloatedness.
>






From aa8vb at yahoo.com  Tue Nov  2 07:34:22 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Tue, 2 Nov 1999 07:34:22 -0500
Subject: __str__ vs. __repr__
In-Reply-To: <000701bf24ef$e2d29b20$0f2d153f@tim>
References: <1270634443-26149069@hypernet.com> <000701bf24ef$e2d29b20$0f2d153f@tim>
Message-ID: <19991102073422.B134170008@vislab.epa.gov>

Tim Peters:
 |[Randall Hopper gives an example]
 |>    class c:
 |>        def __str__(self):  return 'foo'
 |>        def __repr__(self): return 'bar'
 |>
 |>    >>> str(c())
 |>    'foo'
 |>
 |>    >>> str([c(), c(), c()])
 |>    '[bar, bar, bar]'
 |
 |[which Gordon McMillan explains, adding]
 |> Sick enough, but not convoluted enough, to be a Tim-ism...
 |
 |The real illness is that lists (and dicts, and tuples) don't pass
 |str-vs-repr'ness *down*.  That is, even though the example explicitly
 |asks for str of a list, the list object asks for the repr of its
 |elements.  That's convoluted: if (as happens to be true) str is meant to
 |give a friendly string, why do the builtin container types ask their
 |containees to produce unfriendly strings regardless? 

Thanks.  That's the source of my confusion.  I wanted to follow the
language convention.  But Python's internals currently don't follow it, so
the point is somewhat moot.

For consistency, would it make sense to change this for Python 1.5.3 (that
is, have sequence and dict types pass 'str-vs-repr'ness down)?

Thanks,

Randall

-- 
Randall Hopper
aa8vb at yahoo.com



From mhammond at skippinet.com.au  Mon Nov  8 17:45:54 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Mon, 08 Nov 1999 22:45:54 GMT
Subject: .dll and .exe file version informaton
References: <806tck$5o9$1@nnrp1.deja.com>
Message-ID: <CoIV3.29973$pf4.26731@news-server.bigpond.net.au>

Thw win32 extensions currently dont expose that - but someone is working
currently on a win32ver module...  So no for now, but soon...

Mark.

emuller at painewebber.com wrote in message <806tck$5o9$1 at nnrp1.deja.com>...
>Can anyone give me an idea how to quikly get the version info on .dll
>and .exe files under windows?
>
>
>Sent via Deja.com http://www.deja.com/
>Before you buy.





From tbryan at python.net  Wed Nov 24 06:19:30 1999
From: tbryan at python.net (Thomas A. Bryan)
Date: Wed, 24 Nov 1999 06:19:30 -0500
Subject: Python complaints
References: <XFMail.991124113221.mikael@isy.liu.se>
Message-ID: <383BC9C2.901A2D6@python.net>

Mikael Olofsson wrote:
> 
> On 24-Nov-99 Thomas Hamelryck wrote:
>  >  Block delimitation by indenting. 
>
> I am one of those who appreciates this. My reason is the following:

I generally don't mind this aspect of Python, but I've recently found that 
it can cause problems when transmitting code.  For example, I saw some 
code that I wanted to copy in DejaNews c.l.py archives, but the lines 
wrap in narrow columns in DejaNews.  I had to reindent the code.  Grrr.

I have also had problems when e-mailing code when I e-mail the code as 
text in the message instead of attaching a document.  

Is there a better way to transmit code so that possible indentation
munging won't force the reader to reindent? 

vaguely-remembering-something-about-option-BEGIN-and-END-ly yours
---Tom



From kuncej.CENTRAL2.MDCEMAIL at mail.conservation.state.mo.us  Fri Nov 19 11:29:40 1999
From: kuncej.CENTRAL2.MDCEMAIL at mail.conservation.state.mo.us (Jeffrey Kunce)
Date: Fri, 19 Nov 1999 10:29:40 -0600
Subject: UserDict, UserList... UserFile?
Message-ID: <s83526e0.067@mail.conservation.state.mo.us>

>why isn't there a UserFile class, while there is a UserList and a UserDict?
>...
>Otherwise, I'll write it myself; it shouldn't be too hard!

If you write your own, StringIO.py in the standard library might be
a good template to work from. 

  --Jeff




From nospam at bitbucket.com  Sat Nov 27 02:12:29 1999
From: nospam at bitbucket.com (Phil Mayes)
Date: Fri, 26 Nov 1999 23:12:29 -0800
Subject: threading and timeouts
References: <943504970.076.346131773@cs.usyd.edu.au>
Message-ID: <KvL%3.59$E3.4638@newsfeed.avtel.net>

Piers Lauder wrote in message <943504970.076.346131773 at cs.usyd.edu.au>...
>Are timeouts possible inside threads?
>
>Eg: escaping from an attempt to use smtplib.SMTP().sendmail() inside a
>child thread if the remote host is being extremely slow.
>
>It doesn't seem possible to use the traditional signal(SIG_ALRM,...)
>approach because signals only work inside the main thread.
>
>The only option I can think of is to re-write the smtplib module to use
>`select', but does anyone know of a simpler solution?


I struggled with this for some time, failed to get plain old timeouts
working, and ended up with a timeout_socket.py that uses select,
based on code from Lloyd Zusman <ljz at asfast.com>.  I then take advantage
of smtplib.py and poplib.py by inheriting:

class timeout_POP3(poplib.POP3):
    def __init__(self, host, port = poplib.POP3_PORT):
        self.host = host
        self.port = port
        self.sock = timeout_socket.timeout_socket()
        self.sock.connect(self.host, self.port)
        self.file = self.sock.makefile('rb') # same as poplib.POP3
        self._debugging = 0
        self.welcome = self._getresp()

class timeout_SMTP(smtplib.SMTP):
    def connect(self, host='localhost', port = 0):
        """Connect to a host on a given port.
        Override the std SMTP connect in order to use a timeout_socket.
        """
        if not port:
            i = string.find(host, ':')
            if i >= 0:
                host, port = host[:i], host[i+1:]
                try: port = string.atoi(port)
                except string.atoi_error:
                    raise socket.error, "nonnumeric port"
        if not port: port = smtplib.SMTP_PORT
        self.sock = timeout_socket.timeout_socket()
        if self.debuglevel > 0: print 'connect:', (host, port)
        self.sock.connect(host, port)
        (code,msg)=self.getreply()
        if self.debuglevel >0 : print "connect:", msg
        return (code,msg)

This allows me to use poplib.py & smtplib.py unchanged.

The timeout_socket class follows.  I run on Win9x and have a vague
feeling that this won't run on Unix which is why I didn't publish it
before.  If you find out why, yell!

------------ code start ----------------------
""" timeout socket class for connections that can potentially cause
    the server to hang
"""
import socket
import errno
import select
import string
#import appstate  # see comment below

_TIMEOUT = 20.0

class Timeout(Exception):
    pass

class Closing(Exception):
    pass

class timeout_socket:
    def __init__(self, timeout=_TIMEOUT, s=None):
        self.timeout = timeout
        self.inbuf = ''
        if s == None:
            s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
        self.s = s
        self.s.setblocking(0)

    # destructor notes: socket.socket will close when destroyed

    def connect(self, host, port):
        timeout = self.timeout
        s = self.s
        try:
            # Non-blocking mode
            s.setblocking(0)
            s.connect(host,port)
            s.setblocking(timeout != 0)
            return 1
        except socket.error,why:
            # The exception arguments can be (string) or (int,string)
            if len(why.args) == 1:
                code = 0
            else:
                code,why = why
            if timeout:
                s.setblocking(1)
                if code in (errno.EINPROGRESS, errno.EALREADY,
errno.EWOULDBLOCK):
                    # Ok, then wait...
                    r,w,e = select.select([],[s],[],timeout)
                    if w:
                        try:
                            s.connect(host, port)
                            return 1
                        except socket.error,why:
                            # This can throw string or (int,string)
                            if len(why.args) == 1:
                                code = 0
                            else:
                                code,why = why
                            if code == errno.EISCONN:
                                return 1
        except:
            code = 0
            why = 'Unknown error trying to connect'

        # format the error message
        if code:
            code = 'error %d = ' % code
        else:
            code = ''
        msg = 'Connect to %s timed out after %d sec (%s%s)' % (host,
int(self.timeout), code, why)
        raise Timeout, msg

    def send(self, data, timeout=0):
        next = 0
        t = timeout or self.timeout
        total = len(data)
        while 1:
            if appstate.state > appstate.running:
                raise Closing, 'closing'
            r,w,e = select.select([],[self.s], [], t)
            if w:
                buf = data[next:next+8192]
                sent = self.s.send(buf)
                next = next + sent
                if next == total:
                    return
            else:
                raise Timeout, 'timeout while sending "%.20s...": %d sec' %
(data, int(t))

    def recv(self, amt, timeout=0):
        t = timeout or self.timeout
        r,w,e = select.select([self.s], [], [], t)
        if r:
            recvd = self.s.recv(amt)
            return recvd
        if not hasattr(self,'s'): print 'timeout_socket has no socket s
?!',self.s #TMP - Sam's problem 9/21
        raise Timeout, "timeout while receiving from %s: %d sec" %
(`self.s.getpeername()`, int(timeout))

    def recvpending(self, timeout=0):
        """ returns 1/0 """
        return [] != select.select([self.s], [], [], timeout or
self.timeout)[0]

    def read(self, amt):
        """ This only returns when amt has been read or socket times out """
        # new version works with readline (except this doesn't convert CRLF
=> LF)
        while 1:
            if appstate.state > appstate.running:
                raise Closing, 'closing'
            if len(self.inbuf) >= amt:
                break
            self.inbuf = self.inbuf + self.recv(4096)
        data = self.inbuf[:amt]
        self.inbuf = self.inbuf[amt:]
        return data

    # new readline buffers data
    # this -isn't- called for the makefile clone
    def readline(self):
        while 1:
# useful but non-generic code:
#            if appstate.state > appstate.running:
#                raise Closing, 'closing'
            crlf = string.find(self.inbuf, '\r\n')
            if crlf >= 0:
                break
            self.inbuf = self.inbuf + self.recv(4096)
        data = self.inbuf[:crlf] + '\n'
        self.inbuf = self.inbuf[crlf+2:]
        return data

    def close(self):
        self.s.close()

    def makefile(self, flags):
        self.s.setblocking(1) # s is a socket._socketobject
        save = self.s._sock   # stash the real thing
        self.s._sock = self   # replace it by us
        # this call creates a _fileobject that has been passed self as the
socket
        x = self.s.makefile(flags)
        self.s._sock = save   # restore the real socket
        return x

------------ code end ------------------------
--
Phil Mayes    pmayes AT olivebr DOT com
Olive Branch Software - home of Arranger
http://www.olivebr.com/






From jeremy at cnri.reston.va.us  Tue Nov  2 14:40:52 1999
From: jeremy at cnri.reston.va.us (Jeremy Hylton)
Date: Tue, 2 Nov 1999 14:40:52 -0500 (EST)
Subject: Assorted Questions regarding the future of Python...
In-Reply-To: <oqogddnep5.fsf@titan.progiciels-bpi.ca>
References: <6D8A17398E28D3119F860090274DD7DB49833A@pces.cadlab.it>
	<006201bf2517$50edf1f0$f29b12c2@secret.pythonware.com>
	<oqogddnep5.fsf@titan.progiciels-bpi.ca>
Message-ID: <14367.15940.373465.105829@goon.cnri.reston.va.us>

>>>>> "FP" == Fran?ois Pinard <pinard at iro.umontreal.ca> writes:

  FP> I'm curious.  Does Python use standard optimisation techniques
  FP> on that byte code (CSE, strength reduction, and all that)?  I
  FP> would be tempted to guess that the stack machine receives code
  FP> that pretty closely resemble the source code.  Would the stack

Good guess!

  FP> machine be able to express optimisations?  For example, could we
  FP> copy or exchange arbitrary stack elements?

Take a look at the Optimizing Python session at the last Python
conference. 

http://www.foretec.com/python/workshops/1998-11/proceedings.html

In particular, see Skip's paper.

http://www.foretec.com/python/workshops/1998-11/proceedings/papers/montanaro/montanaro.html

Jeremy




From robin at jessikat.demon.co.uk  Sun Nov 28 18:58:54 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Sun, 28 Nov 1999 23:58:54 +0000
Subject: http://www.python.org/ blank ?
Message-ID: <HJYc$DA+GcQ4EwYP@jessikat.demon.co.uk>

for some reason I see http://www.python.org/ as a blank page. Any ideas?
-- 
Robin Becker



From garabik at melkor.dnp.fmph.uniba.sk.spam  Fri Nov  5 09:07:52 1999
From: garabik at melkor.dnp.fmph.uniba.sk.spam (Radovan Garabik)
Date: 5 Nov 1999 14:07:52 GMT
Subject: finding out console user (Win95)
Message-ID: <941814380.235290@cdwork.cvt.stuba.sk>

I am porting my script from unix to windows, and I need to find 
console user under win95 (under WinNT I can use os.environ['USERNAME'],
which is not exactly console user, but good enough for me).
Something like cuserid() or getlogin() as implemented in
Cygwin would be ideal....
Is there a simple way to find it?

TIA

-- 
 -----------------------------------------------------------
| Radovan Garabik http://melkor.dnp.fmph.uniba.sk/~garabik/ |
| __..--^^^--..__         garabik @ fmph.uniba.sk           |
 -----------------------------------------------------------
Antivirus alert: file .signature infected by signature virus.
Hi! I'm a signature virus! Copy me into your signature file to help me spread!



From fredrik at pythonware.com  Tue Nov 30 08:59:23 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 30 Nov 1999 14:59:23 +0100
Subject: Htmlizing text
References: <943909426snz@vision25.demon.co.uk> <19991129170705.A314@better.net> <19991130141224.A2813@stopcontact.palga.uucp> <002301bf3b39$5b5bc9e0$f29b12c2@secret.pythonware.com>
Message-ID: <005b01bf3b3b$1c14d5e0$f29b12c2@secret.pythonware.com>

I wrote:

> Gerrit Holl <gerrit.holl at pobox.com> wrote:
> > By the way, why doesn't the cgi module does this?
> 
> if you use ISO Latin 1 for your pages (which happens to
> be default in HTTP/1.1, HTML 3.0, and real life), you
> don't really have to quote all possible entities just be-
> cause you can...

and yes, your code assumes that the original string
always uses ISO Latin 1.  the cgi version works fine
with plain ASCII, and *any* 8-bit ISO encoding.

</F>




From versation at home.com  Wed Nov  3 10:52:32 1999
From: versation at home.com (Pat Ward)
Date: Wed, 03 Nov 1999 15:52:32 GMT
Subject: working Tcl/Tk8.2.1 & python - broken Tkinter
Message-ID: <4TYT3.7612$OS2.690403@news.rdc1.il.home.com>

I have a working Tcl/Tk8.2.1 & python1.5.2 on RH linux/alpha.

Tkinter passes:
import _tkinter
import Tkinter

fails:
Tkinter._test()
dumping core

I compiled Tcl/Tk with the Compaq beta compiler which gave me a better
build than egcs

I've tried compiling Python with both the Compaq and the egcs. Either gives
me a working _tkinter and a core dumping "test".

I believe that the Modules/Setup is correctly modified and re-made.

I'm baffled. Any ideas?
--
Pat Ward
Versation Corp. - phone 847.506.9095
digital audio and digital media services
425 N. Douglas Avenue
Arlington Heights IL 60004-6122 | http://versation.net/



From hildeb at www.stahl.bau.tu-bs.de  Sat Nov 27 12:35:42 1999
From: hildeb at www.stahl.bau.tu-bs.de (Ralf Hildebrandt)
Date: 27 Nov 1999 17:35:42 GMT
Subject: fcntl/ioctl and O_NONBLOCK
References: <slrn840205.sn4.hildeb@stahlw06.stahl.bau.tu-bs.de> <slrn83t5mv.bhk.hildeb@stahlw06.stahl.bau.tu-bs.de> <81ml5t$eer$0@216.39.151.169> <81p32a$m14$1@216.39.151.169>
Message-ID: <slrn8405je.q0.hildeb@stahlw06.stahl.bau.tu-bs.de>

On 27 Nov 1999 17:06:18 GMT, Donn Cave <donn at oz.net> wrote:

>Right - if there's anything to read, you will read only that and
>not block.  (So of course you must always check how much you actually
>have read!)  Select tells you precisely that your I/O won't block,
>either because there's data to read or end of file.

Fine. My problem here is the following: I start a program (plot) and
feed it with random input data until it crashes. Output is irrelevant
to me, I need only know if it crashes or exit()'s.

On the shell:
% plot < t1 > /dev/null
Segmentation fault.

In my Python program, though: The program ran for a while and never
returned. It didn't crash, it didn't exit, nothing. It just sat there
status "sleep" (displayed top). It didn't consume cpu-time, either.

I figured I should empty fromchild (since the command above generates
8MB of output) from time to time. Then I came across the select()
problem. 

This is also fixed now, but now the popen3 Object generates about 1 MB
of output (which I read using select() and read()), and then it STILL
doesn't return. No crash, no exit. And once more, in top it says
"sleep" once more. 
Any ideas what's going on there?

>I guess - except of course, sockets are bi-directional, and pipes are
>one direction only.

One direction is sufficient for me.



From tim_one at email.msn.com  Tue Nov 30 01:59:52 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Tue, 30 Nov 1999 01:59:52 -0500
Subject: '==' vs. 'is' behavior
In-Reply-To: <oqemd9dk2g.fsf@titan.progiciels-bpi.ca>
Message-ID: <000501bf3b00$802af180$c92d153f@tim>

[Fran?ois Pinard]
> By the way, is there common wisdom (or rather, maybe,
> usage-standards-to-be) about using `is' instead of `=='?

In general, don't -- you'll often end up regretting this cleverness.  Note
that

    a is b

is the same as

    id(a) == id(b)

so the former is *appropriate* only when the latter is *intended*.  A good
example of appropriate use is in copy.deepcopy, where Python's by-value ==
is useless for determining whether a structure contains a cycle.  A terrible
example is "if i is 42":  whether that appears to work is an accident that
may change from release to release or even run to run.

That said, a few objects are guaranteed to exist uniquely, and so it's OK
(but dubious anyway) to use "is" with them.  "x is None", "x is
types.StringType" (and friends), sometimes "x.__class__ is MyClass", and "s
is t" where it's *known* that s and t are both intern()ed strings.

BTW, the truth of "x is y" implies the truth of "x == y", because the first
thing the (current) implementation of comparison does is check for pointer
equality.  However, "x == y" does not imply "x is y".

Clearest and safest:  don't use "is" when "==" would suffice.

Most pragmatic:  don't use "is" unless you know you can get away with it
forever.

Most common:  use "is" randomly, then complain about suspected Python bugs
at 3 in the morning <0.9 wink>.

x-is-y-is-y-is-x-is-not-x-is-not-y-is-not-y-is-not-x-ly y'rs  - tim





From stidolph at origin.ea.com  Tue Nov 16 14:58:45 1999
From: stidolph at origin.ea.com (Stidolph, David)
Date: Tue, 16 Nov 1999 13:58:45 -0600
Subject: list.without()?
Message-ID: <11A17AA2B9EAD111BCEA00A0C9B41793034AAEE7@molach.origin.ea.com>

so ...

def without(source, element):
  while element in source:
    source.remove(element)  

Is that what you are looking for?  It would help if I had read the earlier
posts!  :)

-----Original Message-----
From: Christian Tismer [mailto:tismer at appliedbiometrics.com]
Sent: Tuesday, November 16, 1999 11:15 AM
To: Magnus L. Hetland
Cc: python-list at python.org
Subject: Re: list.without()?




"Magnus L. Hetland" wrote:
> 
> "Mike Fletcher" <mcfletch at vrtelecom.com> writes:
> 
> > Well...
> >
> > def without( source, element):
> >       temp = source[:]
> >       while temp:
> >               try:
> >                       temp.remove( element )
> >               except:
> >                       break
> >       return temp
> 
> Whoa!
> 
> Why on *earth* would you use a while/try/break-combination? IMO a much
> more natural solution would be:
> 
> def without(source, element):
>     result = source[:]
>     if element in result:
>         result.remove(element)
>     return result

But that would be wrong.
"Without" means "there is no such object in it", while list.remove
removes the first occourance of an object.

Mike's implementation appears to be very cheap, since through
the try..except clause, he can combine removal with test
for existance.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home

-- 
http://www.python.org/mailman/listinfo/python-list



From OlaFried at news.vex.net  Sat Nov 13 12:05:35 1999
From: OlaFried at news.vex.net (Ola Fried)
Date: 13 Nov 1999 17:05:35 GMT
Subject: DBF interface
References: <382C38AF.69FFACF4@obop.com.pl>
Message-ID: <80k5ov$2di6$1@hub.org>

On Fri, 12 Nov 1999 16:56:32 +0100, Przemys?aw G. Gawro?ski wrote:
>Is there a DBF file interface module ?

I actually have two of them listed in my little non-dbf database...
http://www.vex.net/~x/parnassus/apyllo.py?find=dbf

The one already mentioned by Michael and S., plus another by Lars Garshol
called dbfreader at http://www.stud.ifi.uio.no/~lmariusg/download/python/

Though if you are on MS platform have ODBC and any MS's (or borland's)
database engine you could access your DBF completely that way probably,
using the odbc module, or the even better mxODBC.

...




From billtut at microsoft.com  Wed Nov 24 04:52:29 1999
From: billtut at microsoft.com (Bill Tutt)
Date: Wed, 24 Nov 1999 01:52:29 -0800
Subject: Python and COM: a bug?
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8100DB915A7@RED-MSG-50>

Well if you can reproduce it, others will try to fix it!
Or at least try to find a workaround, if it can't be fixed. Just sitting
around saying this or that sometimes happens doesn't help many people out.

Saying your C++ code is ok just because it works with VB/VBScript doesn't
necessarily make it ok code. 

> -----Original Message-----
> From: Adrian Eyre [mailto:a.eyre at optichrome.com]
> 
> 
> We did get some oddities running PythonCOM through Rational's 
> Purify: I
> don't have the results any more, but I seem to remember there 
> was internal
> memory corruption somewhere. Not being a COM kind of person, 
> I found an
> easier solution by removing PythonCOM from the project 
> entirely, and going
> via an extension module instead.
> 

Do you recall what kind of scenario caused the problems or did any old
PythonCOM usage cause the problem?

Bill



From mhammond at skippinet.com.au  Tue Nov  2 21:31:44 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 3 Nov 1999 13:31:44 +1100
Subject: Reading .lnk files
In-Reply-To: <004801bf258f$5c19d0d0$0501a8c0@bobcat>
Message-ID: <006301bf25a3$92a7f140$0501a8c0@bobcat>

OK - thanks everyone - it seems certain that the code does indeed
work, and that the problem was something specific to Ivan - quite
possibly his use of the MKS shell.  Ive had positive responses from
NT, 98 and 95 now.

Thanks,

Mark.

[I wrote]
> Ivan has mailed me that this does not work for him (dies with a big
> crash!).  If anyone else has a spare moment, could you please try
this
> code and see if it works?  Whether it does or not, please also
include
> info on the OS you tried it on.  Please mail me personally with this
> info, and I will summarize back to the list.
>
> Thanks,
>
> Mark.
>
> > -----Original Message-----
> > From: python-list-admin at python.org
> > [mailto:python-list-admin at python.org]On Behalf Of Mark Hammond
> > Sent: Wednesday, 3 November 1999 9:44
> > To: 'Ivan Van Laningham'; 'Python Mailing List'
> > Subject: RE: Reading .lnk files
> >
> >
> > Sure is.  For more details, see the COM help file on PyIShellLink.
> >
> > Mark.
> >
> > # DumpShortcut.py
> > # A demo of how to dump a shortcut using Python.
> >
> > # This code dumps the shortcut specificed as sys.argv[1]
> >
> > from win32com.shell import shell
> > import pythoncom
> >
> > def DumpIt(fileName):
> > # Get the shell interface.
> >     sh = pythoncom.CoCreateInstance(shell.CLSID_ShellLink, None, \
> >       pythoncom.CLSCTX_INPROC_SERVER, shell.IID_IShellLink)
> >
> >     # Get an IPersist interface
> >     persist = sh.QueryInterface(pythoncom.IID_IPersistFile)
> >
> >     persist.Load(fileName)
> >
> >     # Get the data
> >     print "Path name:", sh.GetPath(shell.SLGP_SHORTPATH)[0]
> >     print "Args:", sh.GetArguments()
> >     print "Description:", sh.GetDescription()
> >     print "Working directory:", sh.GetWorkingDirectory()
> >
> > if __name__=='__main__':
> >     import sys
> >     DumpIt(sys.argv[1])
> >
> > # demo of usage:
> > C:\>DumpShortcut.py IDLE.lnk
> > Path name: L:\src\PYTHON~1.2\PCbuild\pythonw.exe
> > Args: l:\src\python-cvs\tools\idle\idle.pyw
> > Description:
> > Working directory: L:\src\python-cvs\tools\idle
> >
> >
> > --
> > http://www.python.org/mailman/listinfo/python-list
> >
>
>
> --
> http://www.python.org/mailman/listinfo/python-list
>




From bjorn at roguewave.com  Tue Nov  2 15:59:16 1999
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Tue, 2 Nov 1999 13:59:16 -0700 
Subject: capturing stderr && stdout
Message-ID: <51C9F3C392D0D111A15600A0C99E64A9AF560C@localhost>

What you want is in module popen2, probably popen2.popen3 (they're described
in the module reference).

-- bjorn

> -----Original Message-----
> From: Agent Drek [mailto:drek at monsterbymistake.com]
> Sent: Tuesday, November 02, 1999 12:19 PM
> To: python-list at python.org
> Subject: capturing stderr && stdout
> 
> 
> Hullo,
> 
> I have searched the web and all the python documents/examples 
> I could find
> before posting here. My apologies if I have missed the 
> obvious :) (I bought
> two books as well that I am going through)
> 
> Alot of what I need to do with python depends on launching 
> and controlling
> a subprogram called hscript. It's a 'closed' application that 
> is not a full
> bodied language like python. I am hoping to launch one 
> instance of hscript
> write to it and keep its stdout && stderr.
> 
> Most of our work gets done with csh scripts like the following:
> <csh_script>
> #!/bin/csh -f
> 
> #csh stuff goes here
> 
> hscript -q $HIPDIR/$HIPFILE << ENDCAT >>&! $LOG
> 	# now I write hscript commands
> 	opcf /obj
> 	#etc
> ENDCAT
> exit
> </csh_script>
> 
> here's a stripped down version of what I have thus far.
> 
> <python_script>
> #!/usr/local/bin/python
> 
> import os, sys
> HSCRIPT = "hscript"
> 
> p = os.popen("%s" % HSCRIPT, "w")
> p.write("opcd /obj\n")
> p.write("opadd geo\n")
> p.write("opls\n")
> p.write("quit\n")
> </python_script>
> 
> This is great so far but I just need stderr from hscript 
> (just in case it
> crashes:) and stdout (to keep a neat log file)
> 
> This will need to work on *nix.
> 
> I'm trying to move to python as our main scripting language 
> so that we can
> reuse code and integrate a gui (tkinter) easily into our apps.
> 
> If you're interested in what hscript is you can visit 
http://www.sidefx.com/
and check out houdini.

thanks for any tips,

	=derek

Monster By Mistake Inc > 'digital plumber'
mailto:drek at monsterbymistake.com
http://www.interlog.com/~drek
"This mail was powered by Linux & Vim"


-- 
http://www.python.org/mailman/listinfo/python-list



From phd at phd.russ.ru  Wed Nov 10 08:19:58 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Wed, 10 Nov 1999 13:19:58 +0000 (GMT)
Subject: SWIG and function name question
In-Reply-To: <7vqojf$bgf$1@marcie.fissiontech.com>
Message-ID: <Pine.LNX.4.20.9911101318090.3746-100000@fep132.fep.ru>

On 4 Nov 1999, Scott Barron wrote:
> class Foo:
> 	self.f = foo.__foo  # Create an alias for foo.__foo

> If I comment out self.f = foo.__foo all works fine.  What am I doing wrong
> here?

   Inside a class, all nmaes that starts with double underscore (__)
mangles into Klassname__name.

> 	self.f = foo.__foo  # Create an alias for foo.__foo

   This is equivalent to

> 	self.f = foo.Foo__foo

   And python cannot find any Foo__foo...

Oleg.
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From fdrake at acm.org  Mon Nov 29 17:27:36 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Mon, 29 Nov 1999 17:27:36 -0500 (EST)
Subject: Modules/getpath.c chases links?
In-Reply-To: <199911280746.XAA29926@lotus.CS.Berkeley.EDU>
References: <199911280746.XAA29926@lotus.CS.Berkeley.EDU>
Message-ID: <14402.64984.900304.580726@weyr.cnri.reston.va.us>

Edward Jason Riedy writes:
 > Is there a reason why Modules/getpath.c chases links to get the
 > `correct' path?  I'd like to see that removed, but the comments
 > make me think someone had a specific reason for including it.

Jason,
  Without that, automatic detection of the installation directories
becomes massively fragile.  Consider a fairly reasonable installation:

  $prefix == $exec_prefix == /opt/python/1.5.2

symlinks:

  /usr/local/bin/python --> python152
  /usr/local/bin/python152 --> /opt/python/1.5.2/bin/python

  Without the symlink chasing, the standard library won't be found (I
don't think it even needs to be that complex).

 > I use symlink trees to separate out different versions of packages 
 > distributed by different people (here, stock python 1.5.2 and the
 > LLNL-distributed  NumPy).  The symlink chasing breaks this badly
 > by looking into the symlinked directory rather than using the tree
 > of symlinks.

  I'm not sure I follow.  This sounds incredibly difficult to
maintain.  What does it buy you?

 > Simply removing the code protected by HAVE_READLINK makes my life 
 > much nicer.  I'd really like to see it officially removed.  (If you
 > want a full-featured system for finding the installation, check the
 > kpathsea library in the Web2C distribution of TeX.)

  kpathsea is overkill for most things; anything that doesn't have the 
number of critical configuration items that a TeX installation has
doesn't need such a heavy approach.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From fdrake at acm.org  Fri Nov  5 11:12:50 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Fri, 5 Nov 1999 11:12:50 -0500 (EST)
Subject: Apply a function to each list member?
In-Reply-To: <19991105215504.A497@Ridcully.home>
References: <slrn824gee.1f7.bwinton@tor.dhs.org>
	<XFMail.991105110617.mikael@isy.liu.se>
	<19991105215504.A497@Ridcully.home>
Message-ID: <14371.514.406346.901988@weyr.cnri.reston.va.us>

Malcolm Tredinnick writes:
 > one-to-one, not one-to-many). So some mapping that sends, say, the number 2
 > to the numbers 3 and 4 is not a function (but a mapping that sends both the
 > numbers 3 and 4 to the number 2 is a function, since for each input there

  For people who are confused, this is also called a "relation."  All
functions are relations, but not all relations are functions (for
exactly the reasons Malcolm gives).


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From news at dorb.com  Wed Nov 24 10:47:10 1999
From: news at dorb.com (Darrell)
Date: Wed, 24 Nov 1999 10:47:10 -0500
Subject: CRC-module
References: <38396766.B189D15A@bibsyst.no> <383C1089.1D0F7C18@bibsyst.no>
Message-ID: <gKT_3.2758$X5.264284@typhoon.nyroc.rr.com>

>From Fredrik's post:
    sha (160-bit digest)

So more bits are better and 160 bits is best.
UNTESTED:
shaHash=sha.sha(open(yourFile).read()).digest()

The output is a number. But if you look at any 8bits of the 160 possible
bits.
Some may be printable characters.

--
--Darrell
"Thomas Weholt" <thomas at bibsyst.no> wrote in message
news:383C1089.1D0F7C18 at bibsyst.no...
> Hi,
>
> Ok, so I`ve looked into zlib.crc32 and zlib.adler32. They seem easy
> enough to use, but I thought crc-codes had characters and numbers in
> them, not just a plain integer like the methods above return. ( As you
> can see, I`m a complete ass on this subject, but don`t have time to do
> the proper research myself, and was hoping for a "quick fix" ... )
>
> A friend of mine mentioned that I should try SHA-1 instead, for more
> accuracy. Can anybody give me an example on how to compute crc-codes,
> using zlib or preferrably some more accurate method, for single files ??
>
> If this is all it takes :
>
> crc = module_name.crc_method(file)
>
> and comparison is done like :
>
> if (crc1 == crc2): print "Equal."
> else : print "Different."
>
> then all I need is the name of the most effective/accurate module to
> use.
>
> If, for some strange reason, I should use one module instead of another,
> that info would be interesting too.
>
> I`ll stop bother you now and let you get back to work. ;->
>
> Thomas Weholt
>
> --
> http://www.python.org/mailman/listinfo/python-list





From tekhir at yahoo.com  Sun Nov 14 18:47:46 1999
From: tekhir at yahoo.com (Jeremy Misavage)
Date: Sun, 14 Nov 1999 18:47:46 -0500
Subject: Setting up Pythong for IIS/PWS
Message-ID: <382F4A22.514281A8@yahoo.com>

First I want to thank everyone who answered my previous question.  All of you
helped me out.  Now I have anotehr uestion.

I don't think I setup IIS on NT and PWS on Win95 correctly.  Supposly you just:

"HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\W3SVC\Parameters\ScriptMap 

and enter the following line (making any specific changes that your system may
need) 

.py :REG_SZ: c:\<path to python>\python.exe -u %s %s "

But I keep getting a 405 error "Method not allowed".  I'm assuming this is
because I didn't add the above correctly.  Anyone want to walk me through or
poit me to a site that already has a setup tutorial?  (I'm kinda out of my
element in Widows)



From htrd90 at zepler.org  Sun Nov  7 09:13:45 1999
From: htrd90 at zepler.org (Toby Dickenson)
Date: Sun, 07 Nov 1999 14:13:45 +0000
Subject: __str__ vs. __repr__
References: <199911041325.IAA14001@eric.cnri.reston.va.us> <000001bf275c$02370d20$1d2d153f@tim>
Message-ID: <u4glOIWXzGrqvJnI+IlPAQeurz1J@4ax.com>

"Tim Peters" <tim_one at email.msn.com> wrote:

>Let me back off to what repr and str "should do":
>
>repr(obj) should return a string such that
>    eval(repr(obj)) == obj

I don't understand the motivation for this requirement. Why would
anyone want to pass such a string to eval? If you anticipate the need
for reconstructing the object from a textual representation, then
surely pickle is a better option?

Alternatively, most classes could implement __repr__ as :-)

    def __repr__(self):
        return "pickle.loads(%s)" % repr(pickle.dumps(self))



>repr() currently cheats in another way:  for purely aesthetic reasons,
>repr(float) and repr(complex) don't generate enough digits to allow exact
>reconstruction of their arguments (assuming high-quality float<->string in
>the platform libc).  This is a case where an argument appropriate to str()
>was applied to repr(), not because it makes *sense* for repr, but because so
>much output goes thru an *implicit* repr() now and you didn't want to see
>all those "ugly" long float strings <0.50000000000000079 wink>.

OK, so repr as it stands today is clearly not suitable for formatting
for people, and is less than ideal for formatting for the machine.
However there is a third audience for who it is ideal, the programmer,
who need a representation for use in development tools such as
debuggers, or the interactive mode.

Tim's Rat class is a good example of how __repr__ should not be
implemented, because it exposes the internal representation of the
object, not the external value.

However, using a small number of digits for repr(float) does make
sense because it still provides sufficient precision for most
programmers.



Toby Dickenson



From m.faassen at vet.uu.nl  Fri Nov 19 05:44:08 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 19 Nov 1999 10:44:08 GMT
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <JlFX3.3659$s87.67930@dfiatx1-snr1.gtei.net> <80oqn9$ocq$2@newshost.accu.uu.nl> <oLUX3.286$yA4.4821@dfiatx1-snr1.gtei.net> <80p9su$ppo$2@newshost.accu.uu.nl> <80piru$13rl$1@hub.org> <80re5d$55a$1@newshost.accu.uu.nl> <fjPY3.770$An3.14793@newsfeed.avtel.net> <810jto$lo0$1@newshost.accu.uu.nl> <811pdu$dvo$1@newshost.accu.uu.nl> <8138ek$22mj$1@hub.org>
Message-ID: <8139lo$jc0$1@newshost.accu.uu.nl>

T. wrote:
> On 18 Nov 1999 21:00:46 GMT, Martijn Faassen <m.faassen at vet.uu.nl> wrote:
>>with the same speech patterns _from vex.net_, I meant! deja nor I are that

> Argh! They're on to me. I see i'll have to randomise news servers 
> for posting more, and deja too. (-; Oh yaw, and type with a dutch
> accent! Happy trails.
> ...

Don't underestimate the Python Secret Underground! :)

Now I *still* don't know why. Is it just that you're so dynamic? :)

Wonders-if-he-himself-types-with-a-Dutch-accent-ly yours,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?



From skaller at maxtal.com.au  Thu Nov 25 18:34:29 1999
From: skaller at maxtal.com.au (skaller)
Date: Fri, 26 Nov 1999 10:34:29 +1100
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <81g8po$reo$1@mach.vub.ac.be> <81h2hh$85c$1@nntp9.atl.mindspring.net>
Message-ID: <383DC785.5C81CA06@maxtal.com.au>

Aahz Maruch wrote:
> 
> In article <81g8po$reo$1 at mach.vub.ac.be>,
> Thomas Hamelryck <thamelry at vub.ac.be> wrote:
> >
> >Block delimitation by indenting. This makes python look like a porsche
> >with a caravan attached :-). As far as I know, most python users even
> >seem to like this for some reason.
> 
> I agree with you and I'm also one of the people who likes it.  Here's
> why: it makes Python code *readable*. 

I agree. I'll go further. I like to write code, literate programming
style FOR PUBLICATION. Just try typesetting C for publication
(and I mean ALL the required code, not just snippets).

Compare with Python. you don't even need to understand python
to read an algorithm written using it:

	for i in [1,2,3]:
		print i, i * i

is plain obvious 'psuedo code' that has the nice
feature that it can actually be executed.

-- 
John Skaller, mailto:skaller at maxtal.com.au
10/1 Toxteth Rd Glebe NSW 2037 Australia
homepage: http://www.maxtal.com.au/~skaller
voice: 61-2-9660-0850



From tbryan at python.net  Fri Nov 12 18:31:06 1999
From: tbryan at python.net (Thomas A. Bryan)
Date: Fri, 12 Nov 1999 18:31:06 -0500
Subject: making the Python case to my dept.
References: <9911051740130D.02530@william.jinsa.org> <3823FC43.CF7FBD71@python.net> <99110609085702.02970@quadra.teleo.net> <382467FA.577E70A0@callware.com> <3824EF76.113E3699@python.net>
Message-ID: <382CA33A.89B50E87@python.net>

"Thomas A. Bryan" wrote:
> 
> Ivan Van Laningham wrote:
> 
> > > Since this is a FAQ, I'd vote for preserving Thomas' answer somewhere
> > > on the python.org Web site, so people could point to the URL when the
> > > question comes up again.

It's now in the FAQ under section 2.

---Tom



From mlh at vier.idi.ntnu.no  Mon Nov  8 09:55:08 1999
From: mlh at vier.idi.ntnu.no (Magnus L. Hetland)
Date: 08 Nov 1999 15:55:08 +0100
Subject: What are Aspects? (Re: How can I make my assertions smarter?)
References: <Pine.WNT.4.05.9911032042440.109-100000@david.ski.org> <slrn824lqc.ngk.wtanksle@hawking.armored.net> <7vvq63$ehm$1@nntp4.atl.mindspring.net> <3826C32A.FB47EF6@compaq.com>
Message-ID: <y0jpuxlnicz.fsf@vier.idi.ntnu.no>

Greg Ewing <greg.ewing at compaq.com> writes:

> Smacks of the legendary
> come-from statement, somehow...

Sounds cool... Only a myth?

> 
> Greg

--

  Magnus          Echelon jamming noise:
  Lie             FBI CIA NSA Handgun Assault Bomb Drug Terrorism
  Hetland         Special Forces Delta Force AK47 Hillary Clinton 



From ivanlan at callware.com  Tue Nov  2 15:14:22 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Tue, 02 Nov 1999 13:14:22 -0700
Subject: Barf
References: <381F37FA.3899AAEF@callware.com>
Message-ID: <381F461E.3E136341@callware.com>

Hi All--

Ivan Van Laningham wrote:
> 

[snip]

> 
> It works fine, but if I move the mouse pointer out of the info box
> (select Help->About) and back to the main window, I get a segmentation
> violation.
> 

If I add a Frame to the root window, it works perfectly.

Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From steve at acer.spvi.com  Wed Nov  3 06:24:43 1999
From: steve at acer.spvi.com (Steve Spicklemire)
Date: Wed, 3 Nov 1999 06:24:43 -0500 (EST)
Subject: Python director XTra....  cute 'robot' toy for mac *and* windows....
In-Reply-To: <199911022220.RAA21086@acer.spvi.com> (message from Steve
	Spicklemire on Tue, 2 Nov 1999 17:20:28 -0500 (EST))
References: <199910291511.KAA26339@acer.spvi.com> <199911022220.RAA21086@acer.spvi.com>
Message-ID: <199911031124.GAA24189@acer.spvi.com>

Windows version now available.... 

There is now a Windows version of the Python Xtra for Macromedia
director. There are also a couple of dumb bug fixes. To use the Xtra
you'll need to have Python installed on the system where the movie (or
projector) is running. (Doesn't everyone have python installed?)

find it at:

http://opensource.spvi.com

Source code is, of course, included in this release. Any help would be
most welcome. ;-) If you have any comments or problems, please email
me directly since it's totally impossible for me to keep up with this 
list and get anything else done.

thanks!
-steve




From fuming at venus.radsci.uci.edu  Fri Nov  5 10:34:25 1999
From: fuming at venus.radsci.uci.edu (Fuming Wang)
Date: Fri, 5 Nov 1999 07:34:25 -0800
Subject: Problems with Tkinter while the thread module is configured
In-Reply-To: <m3eme59dg1.fsf@atrus.jesus.cam.ac.uk>
References: <slrn824gee.1f7.bwinton@tor.dhs.org> <XFMail.991105110617.mikael@isy.liu.se> <19991105215504.A497@Ridcully.home> <14370.53219.667492.946075@gargle.gargle.HOWL> <27NwjKAK1uI4EwdJ@jessikat.demon.co.uk> <m3eme59dg1.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <Pine.SOL.4.05.9911050727040.24818-100000@venus.radsci.uci.edu>

Hi,

Is there anyone having this problem too? When "thread" module is
configured, Tkinter programs would exit right after start.

Environment:
	Python1.5.2
	LinuxThread-0.71
	Libc5.4.46
	Linux2.0.33

Fuming Wang







From bellman at lysator.liu.se  Wed Nov 17 07:48:01 1999
From: bellman at lysator.liu.se (Thomas Bellman)
Date: 17 Nov 1999 12:48:01 GMT
Subject: cgi.py inefficiency (with fix)
Message-ID: <80u861$ngs$1@newsy.ifm.liu.se>

I was profiling an embedded web server we have written in Python,
and found that it spent a lot of time in FieldStorage.has_key()
and FieldStorage.__getitem__() from cgi.py.  So I looked at the
code for the FieldStorage class, and found that it basically
emulates a normal dictionary, but is implemented using a linear
list of fields.  Thus, has_key() and __getitem__() must do linear
searches to find things.  Is there any reason why this is done,
instead of having a dictionary of lists?

I did a quick stab at replacing the list with a dictionary, and
in our application, we got a threefold increase in speed per
request...  Of course, in our application we have forms sometimes
containing well over a thousand entries.  It is probably not very
common for people to have such large forms, but it is a pity that
the cgi module does not scale well.

Anyway, here is a diff of the changes I made.  I haven't tested
it very extensively (in our application we have subclassed the
FieldStorage class, and use a modified __init__() method, which
makes it difficult to fully test the real FieldStorage class),
but from what I have tested, it seems to work.

I started with revision 1.45 of cgi.py, which I checked out from
the CVS archives this Monday.


===================================================================
RCS file: /projects/cvsroot/python/dist/src/Lib/cgi.py,v
retrieving revision 1.45
diff -u -5 -r1.45 cgi.py
--- cgi.py	1999/06/11 18:26:09	1.45
+++ cgi.py	1999/11/17 12:30:23
@@ -724,11 +724,11 @@
 
     """Like FieldStorage, for use when no file uploads are possible."""
 
     # Dummy attributes
     filename = None
-    list = None
+    dict = None
     type = None
     file = None
     type_options = {}
     disposition = None
     disposition_options = {}
@@ -890,11 +890,12 @@
                 pass
             if maxlen and clen > maxlen:
                 raise ValueError, 'Maximum content length exceeded'
         self.length = clen
 
-        self.list = self.file = None
+        self.file = None
+	self.dict = None
         self.done = 0
         self.lines = []
         if ctype == 'application/x-www-form-urlencoded':
             self.read_urlencoded()
         elif ctype[:10] == 'multipart/':
@@ -912,74 +913,69 @@
             raise AttributeError, name
         if self.file:
             self.file.seek(0)
             value = self.file.read()
             self.file.seek(0)
-        elif self.list is not None:
-            value = self.list
+        elif self.dict is not None:
+            value = self.dict.values()
         else:
             value = None
         return value
 
     def __getitem__(self, key):
         """Dictionary style indexing."""
-        if self.list is None:
+        if self.dict is None:
             raise TypeError, "not indexable"
-        found = []
-        for item in self.list:
-            if item.name == key: found.append(item)
-        if not found:
-            raise KeyError, key
+        found = self.dict[key]
         if len(found) == 1:
             return found[0]
         else:
             return found
 
     def keys(self):
         """Dictionary style keys() method."""
-        if self.list is None:
+        if self.dict is None:
             raise TypeError, "not indexable"
-        keys = []
-        for item in self.list:
-            if item.name not in keys: keys.append(item.name)
-        return keys
+        return self.dict.keys()
 
     def has_key(self, key):
         """Dictionary style has_key() method."""
-        if self.list is None:
+        if self.dict is None:
             raise TypeError, "not indexable"
-        for item in self.list:
-            if item.name == key: return 1
-        return 0
+        return self.dict.has_key(key)
 
     def __len__(self):
         """Dictionary style len(x) support."""
-        return len(self.keys())
+        return len(self.dict)
 
     def read_urlencoded(self):
         """Internal: read data in query string format."""
         qs = self.fp.read(self.length)
-        self.list = list = []
+        self.dict = dict = {}
         for key, value in parse_qsl(qs, self.keep_blank_values,
                                     self.strict_parsing):
-            list.append(MiniFieldStorage(key, value))
+	    l = dict.get(key, [])
+	    l.append(MiniFieldStorage(key, value))
+            dict[key] = l
         self.skip_lines()
 
     FieldStorageClass = None
 
     def read_multi(self, environ, keep_blank_values, strict_parsing):
         """Internal: read a part that is itself multipart."""
-        self.list = []
+        self.dict = {}
         klass = self.FieldStorageClass or self.__class__
         part = klass(self.fp, {}, self.innerboundary,
                      environ, keep_blank_values, strict_parsing)
         # Throw first part away
         while not part.done:
             headers = rfc822.Message(self.fp)
             part = klass(self.fp, headers, self.innerboundary,
                          environ, keep_blank_values, strict_parsing)
-            self.list.append(part)
+	    l = self.dict.get(part.name, [])
+	    l.append(part)
+            self.dict[part.name] = l
         self.skip_lines()
 
     def read_single(self):
         """Internal: read an atomic part."""
         if self.length >= 0:
===================================================================

-- 
Thomas Bellman,   Lysator Computer Club,   Link?ping University,  Sweden
"Beware of bugs in the above code; I have    !  bellman @ lysator.liu.se
 only proved it correct, not tried it."      !  Make Love -- Nicht Wahr!



From gerrit.holl at pobox.com  Fri Nov 19 08:01:08 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Fri, 19 Nov 1999 14:01:08 +0100
Subject: Beginning programming with Python
In-Reply-To: <assonant-80s673/INN-2.2.1/basilar@broadway.news.is-europe.net>
References: <sng0ychgwt.fsf@motorola.com> <382B3A53.5816B2@callware.com> <382B3D79.74CD3E0F@email.sps.mot.com> <80p9ds$ppo$1@newshost.accu.uu.nl> <asteroidal-80psfo/INN-2.2.1/bat@broadway.news.is-europe.net> <38312D55.A36EB56D@iqsoft.hu> <antigorite-80rol2/INN-2.2.1/attentive@broadway.news.is-europe.net> <19991116173142.A31379@optiplex.palga.uucp> <assonant-80s673/INN-2.2.1/basilar@broadway.news.is-europe.net>
Message-ID: <19991119140108.C4306@optiplex.palga.uucp>

Janos Blazi wrote:
> > the less time it costs to program it, the more time it costs to run it,
> > the more time it costs to program it, the less time it costs to run it.
> 
> Is this some law of nature I missed at university?

Well, no. I'm 14 :-)

> But obviously, as far as raw speed is concerned Python cannot compete with
> C. Nobody can attempt to deny that.

Yes

> So if speed is an issue for you, use C.

But if *development* speed is an issue, use Python.

> Or even assembly language.
> If you do not like Python, use C anyway. Why not? Actually I do not care.

I like Python. Some people don't like Python *because* of it's easy syntax,
but I find that hard to believe.

regards,
Gerrit.

-- 
"We all know Linux is great...it does infinite loops in 5 seconds."
(Linus Torvalds about the superiority of Linux on the Amterdam
Linux Symposium)



From gmcm at hypernet.com  Tue Nov 16 16:00:23 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 16 Nov 1999 16:00:23 -0500
Subject: Python in business programming [was Re: Beginning programming with Python]
In-Reply-To: <383195C1.826B3342@iqsoft.hu>
Message-ID: <1269336472-10033930@hypernet.com>

Nemeth Miklos writes:

[snip]
> The only thing I wanted, is to point out a potential shortcoming (if at all) of
> Python -- I am not sure if the lack of compile-time type safety is really a
> badly needed feature.

It is, but only if you're used to relying on it. That is, some 
people miss it dearly. Coming from a background of languages 
that did static type checking, I did at first. Adapting has meant 
big changes to my development style (the code, compile, test 
cycle is now rarely more than half an hour). Test suites are 
more important (but also easier). For large systems, 
decomposing the problem cleanly is very important - to 
preserve your sanity you need the smallest, cleanest 
interfaces possible between subsystems. But, if at first you 
foul it up, you can still rewrite most of it before lunch.

> I am neither a physicist nor a mathematician, I am a simple business software
> engineer. I like Python but I'd like to be convinced that Python is not only
> for scientific programming, but also for business programming. 

I can assure you that there are lots of business programmers 
who use and love Python. Probably around half of c.l.py 
contributers are involved in business, not math / science. 
Businesses tend to be pretty conservative, however, and 
exercise rigid control over the tools their programmers use. So 
acceptance at the mangerial level is slow.

In my business as a consultant, most of my clients want C, 
C++ or Java. That doesn't mean that I don't write it (and rewrite 
it until it's right) in Python first. And sometimes they accept 
delivery in Python (when they hear what the price tag would be 
to get it in C, for example).

- Gordon



From tbryan at python.net  Fri Nov 12 18:16:50 1999
From: tbryan at python.net (Thomas A. Bryan)
Date: Fri, 12 Nov 1999 18:16:50 -0500
Subject: Where are Python's IDEs documented?
Message-ID: <382C9FE1.944D3494@python.net>

I wanted to point users to some documentation about IDEs for Python:
PythonWin, IDLE, Emacs, etc.  I was hoping to find one page on the 
Python website that at least listed all of the options, maybe with 
screenshots.  

I searched around http://www.python.org for a while, but I never found 
such a page.  Is there one?  I think that such a page would be valuable, 
especially for first-time programmers.

---Tom



From torkristensen at my-deja.com  Tue Nov 16 05:22:37 1999
From: torkristensen at my-deja.com (torkristensen at my-deja.com)
Date: Tue, 16 Nov 1999 10:22:37 GMT
Subject: Python accessing Webcam?
Message-ID: <80rb9c$m87$1@nnrp1.deja.com>

Hi all,

I was wondering if anyone could point me in a
direction that would let me grab a frame from a
webcam.

Is there a way to access DEVICE.framebuffer?

thanks in advance
t


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



From gvwilson at nevex.com  Sun Nov 28 13:47:51 1999
From: gvwilson at nevex.com (Greg Wilson)
Date: Sun, 28 Nov 1999 13:47:51 -0500
Subject: discrete event simulation?
Message-ID: <xOe04.8702$cz3.315201@nnrp1.uunet.ca>

Has anyone written a discrete event simulation framework in Python?  If so,
I'd be grateful for pointers.
Thanks,
Greg






From prodos8 at yahoo.com  Tue Nov 30 18:38:28 1999
From: prodos8 at yahoo.com (help)
Date: Tue, 30 Nov 1999 18:38:28 -0500
Subject: For sale: "Learning Python" + "Python Pocket Ref."
Message-ID: <301119991838284235%prodos8@yahoo.com>

$18.00 including UPS ground shipping anywhere in the US:

I have for sale a new, almost unused copy of the ORA book Learning
Python and the supplement Python Pocket Reference.  Both are good,
well-done books that have gotten great reviews.  Learning Python is
$23.96 new from amazon.com, plus shipping; the pocket reference is
$5.56.  You save, by buying my copy for $18, $11 plus shipping.  email
me at brian at innobyte.com if you are interesed.  If you are in the
Boston area and can pick it up, they are yours for $15.00.  thanks.



From dnagata at creo.com  Fri Nov 12 23:00:39 1999
From: dnagata at creo.com (Dale Nagata)
Date: Fri, 12 Nov 1999 20:00:39 -0800
Subject: print "foo", without a space
References: <382AD3BB.69899CBD@helsinki.xx>
Message-ID: <382CE267.5A94@creo.com>

Mikael Johansson wrote:
> Is there another way of getting print to not make a newline than
> using:
>     print "foo",  ?
> This because the comma always produces a space

[snip]

have you tried

	sys.stdout.write( "foo" )


--
Dale Nagata         |  tel : +1 604.451.2700 ext. 2254 (UTC-0800)
Software Developer  |  fax : +1 604.437.9891 
Creo Products Inc.  |  pgr : +1 604.691.8279
Burnaby BC Canada   |  http://www.creo.com/



From pinard at iro.umontreal.ca  Tue Nov  2 13:22:10 1999
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 02 Nov 1999 13:22:10 -0500
Subject: Assorted Questions regarding the future of Python...
In-Reply-To: Skip Montanaro's message of "Tue, 2 Nov 1999 13:19:02 -0600 (CST)"
References: <6D8A17398E28D3119F860090274DD7DB49833A@pces.cadlab.it> <006201bf2517$50edf1f0$f29b12c2@secret.pythonware.com> <oqogddnep5.fsf@titan.progiciels-bpi.ca> <14367.14630.983098.834929@dolphin.mojam.com>
Message-ID: <oqg0yooist.fsf@titan.progiciels-bpi.ca>

Skip Montanaro <skip at mojam.com> ?crit:

> For pointers to [optimisation related] stuff [...]

Thanks a great deal, Skip, for all these pointers, I'll preciously save them.
Who knows, maybe that one of these days, I'll find some more time and try to
go play into your sandbox, if there is some corner for me at that time ;-).

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





From ivanlan at callware.com  Sat Nov 27 15:02:20 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Sat, 27 Nov 1999 13:02:20 -0700
Subject: yet 'nother newbie question
References: <3840300F.66DC46DD@wjk.mv.com>
Message-ID: <384038CC.BBB47AE4@callware.com>

Hi All--

"William J. King" wrote:
> 
> Hi:
> 
> When passing an argument to a function etc. -- how does one determine
> that it is
> an integer that is being passed and not a string for example... Is there
> something
> available to determine it's type. The only way I can figure is to use
> 'try' and
> 'except' for a partial solution:
> 
> try:
>     x = x + ""    # suceeds then its a string -- if so
>    x = string.atoi(x)
> except:
>     x = x + 0    #This sort of partially ensures that int is passed.....
> 
> or if you have the reverse situation?...
> 
> Hmmm -- It seems python knows the difference internally, how can I
> determine
> types and work with variables without having to resort to exception
> checking...
> Hope this makes sense... Thanks.....
> 

type(x)

That is, 

if type(x) == type(""):
	print "I'm a string"
elif type(x) == type(0):
	print "I'm an integer"
else:
	print "I'm something else"


type() returns a repr (a printable string) object that might look like
this:

	<type 'string'>

There is a module (types) that you can import which has shortcuts in it
so that you can call

if type("")==StringType:
	blah blah

but I generally don't bother, since it saves almost nothing on typing.

<and-we-all-know-keystrokes-determine-reality>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours
----------------------------------------------



From gerrit.holl at pobox.com  Wed Nov 10 08:11:43 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Wed, 10 Nov 1999 14:11:43 +0100
Subject: MSWIN:getch() UNIX:?
In-Reply-To: <80a3rk$dg4$1@nslave2.tin.it>
References: <80a3rk$dg4$1@nslave2.tin.it>
Message-ID: <19991110141143.A32387@optiplex.palga.uucp>

Francesco wrote:
> There's a module with a getch() function that can run on all platforms?

Write a function what checks on witch OS it is and import the getch() from the
curses module if you're on unix but from another module on dos. For unix, read
the curses howto.

regards,
Gerrit.
-- 
linuxgames.nl.linux.org	 All about games on Linux in Dutch (under construction).
www.nl.linux.org	       The Dutch resource for Dutch Linux documentation.
www.nl.linux.org/discoverb		    Learn foreign words and definitions.
www.nl.linux.org/~gerrit/asperger	 I'm an Asperger, go here for more info.
www.nl.linux.org/~gerrit			    About me and pictures of me.



From sramanath at my-deja.com  Tue Nov  9 16:05:05 1999
From: sramanath at my-deja.com (sramanath at my-deja.com)
Date: Tue, 09 Nov 1999 21:05:05 GMT
Subject: Detect file existance before open?
References: <zWaO3.2925$lv2.105561@news.uswest.net> <380c92a7.412580789@192.77.87.74>
Message-ID: <80a29u$gn9$1@nnrp1.deja.com>

import commands
commands.getstatus('<filename>')


In article <380c92a7.412580789 at 192.77.87.74>,
  jlj at cfdrc.com (Lyle Johnson) wrote:
> os.path.exists(filename)
>
> On Sat, 16 Oct 1999 21:06:07 -0600, "Bill Rodgers" <Bill at nospam.com>
> wrote:
>
> >I would like to check for the existance of a file before I attempt an
open
> >and maybe throw an exception.  I don't see a way to do this.  I would
like
> >to tell the user that the file he specified doesn't exist.  Is there
a way
> >to do this or do I need to catch the exception and deal with it that
way?
> >Thanks,
> >--
> >Bill Rodgers
> >Enterprise Management Consultant
> >Specialty: BMC/Patrol
> >brodg1 at ix.netcom.com
> >
> >
>
>


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



From martian1 at dds.nl  Sun Nov 28 15:24:17 1999
From: martian1 at dds.nl (Martian)
Date: Sun, 28 Nov 1999 21:24:17 +0100
Subject: "Import error"--MySQLmodule broke betw. Python 1.5.1 and 1.5.2
Message-ID: <38418F71.A39C213E@dds.nl>

Hello,

I am programming a CGI app which uses the MySQLmodule 1.4 python interface to
the MySQL database (version 3.22.27).

I've downloaded and installed Python 1.5.2 for my FreeBSD 3.3 box.

In August, I programmed something similar, also using the MySQLmodule,
but then I used Python 1.5.1

The problem is as follows:
When I start Python (interactively) and type:
>>> import MySQLmodule
I get:
ImportError: dynamic module does not define init function (_initMySQL)

I've never seen this problem before; in Py1.5.1 MySQLmodule could be
used just fine, both interactively and in a script.

The MySQLmodule hasn't changed--indeed, hasn't since Oct. 1998

The module itself is written in C, and I compiled it without errors or 
warnings. (Hard though that may seem with a C program! ;-)

I suppose the easy solution would be to simply remove Python 1.5.2 and install
Python 1.5.1, but I would much rather stay with 1.5.2 to stay up-to-date.
(Besides, aren't there bugfixes in this release? Better to have them, too)
Could you tell me if I could perhaps define some kind of "dummy" C function
within the MySQLmodule C code which would satisfy Python--an "__initMySQL__"
function which simply does nothing.

I am inexperienced in programming C (I know the basics, though) and have never
integrated C and Python. If I define an empty __initMySQL__ function in
MySQLmodule.c, will that break the module? Does the hypothetical initMySQL 
function somehow need to register MySQLmodule's objects, or in any other
way do something more than just *exist* ?

Thank you for any help you can offer,

Martijn Herber

P.S Since I am not (yet) subscribed to the list itself, please also respond
to my email adress directly: martian1 at dds.nl



From gerrit.holl at pobox.com  Wed Nov 10 08:13:15 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Wed, 10 Nov 1999 14:13:15 +0100
Subject: __str__ vs. __repr__
In-Reply-To: <000301bf2b46$4d7e6f60$d82d153f@tim>
References: <14376.31199.948463.211294@goon.cnri.reston.va.us> <000301bf2b46$4d7e6f60$d82d153f@tim>
Message-ID: <19991110141315.B32387@optiplex.palga.uucp>

Tim Peters wrote:
> No, in that case it would have been named __representation__.  __repr__ is
> actually an acronym, meaning Regular-Expression Parsable Rendition.
> Similarly __str__ is short for Simpler Than Repr.  I never learned what
> __init__ stands for -- it never interested Tim <wink>.

:-)

> guess-what-wink-stands-for-and-win-a-fabulous-prize!-ly y'rs  - tim

Windows-Is-Not-Kewl?

regards,
Gerrit.
-- 
linuxgames.nl.linux.org	 All about games on Linux in Dutch (under construction).
www.nl.linux.org	       The Dutch resource for Dutch Linux documentation.
www.nl.linux.org/discoverb		    Learn foreign words and definitions.
www.nl.linux.org/~gerrit/asperger	 I'm an Asperger, go here for more info.
www.nl.linux.org/~gerrit			    About me and pictures of me.



From aahz at netcom.com  Thu Nov 18 16:21:13 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 18 Nov 1999 21:21:13 GMT
Subject: Beginning programming with Python
References: <sng0ychgwt.fsf@motorola.com> <antigorite-80rol2/INN-2.2.1/attentive@broadway.news.is-europe.net> <19991116173142.A31379@optiplex.palga.uucp> <assonant-80s673/INN-2.2.1/basilar@broadway.news.is-europe.net>
Message-ID: <811qk9$3b4$1@nntp9.atl.mindspring.net>

In article <assonant-80s673/INN-2.2.1/basilar at broadway.news.is-europe.net>,
Janos Blazi <jblazi at netsurf.de> wrote:
>> the less time it costs to program it, the more time it costs to run it,
>> the more time it costs to program it, the less time it costs to run it.
>
>Is this some law of nature I missed at university?
>But obviously, as far as raw speed is concerned Python cannot compete with
>C. Nobody can attempt to deny that. So if speed is an issue for you, use C.
>Or even assembly language.
>If you do not like Python, use C anyway. Why not? Actually I do not care.

The point is that, generally speaking, programming in Python takes less
time than programming in C, but C programs are faster.  So you trade
development time for code speed.  OTOH, programming in C is enough more
difficult that if your speed requirements are based on finding a good
algorithm, your Python program might well be faster (because you can try
more algorithms per unit time in Python).
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From greg.ewing at compaq.com  Wed Nov  3 03:58:37 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Wed, 03 Nov 1999 21:58:37 +1300
Subject: Printing of individually returned values when running interactively
References: <9t966zk6gvk.fsf@mraz.iskon.hr>
Message-ID: <381FF93D.861C0DBB@compaq.com>

Hrvoje Niksic wrote:
> 
> I've just noticed (the hard way) a very unusual feature of Python's
> interactive interpreter mode.  If you evaluate several values on
> several lines, all the values are printed!
>
>  Why would anyone want a thing like that?

I doubt anyone ever explicitly wanted it. I suspect it's a
side effect of the implementation. I just did the following
experiment:

>>> c = compile("""
if 1:
	1
	2
	3
""", "", "single")
>>> c
<code object ? at 815b00, file "", line 0>
>>> from dis import dis
>>> dis(c)
          0 SET_LINENO          0

          3 SET_LINENO          2
          6 LOAD_CONST          0 (1)
          9 JUMP_IF_FALSE      25 (to 37)
         12 POP_TOP        

         13 SET_LINENO          3
         16 LOAD_CONST          0 (1)
         19 PRINT_EXPR     

         20 SET_LINENO          4
         23 LOAD_CONST          1 (2)
         26 PRINT_EXPR     

         27 SET_LINENO          5
         30 LOAD_CONST          2 (3)
         33 PRINT_EXPR     
         34 JUMP_FORWARD        1 (to 38)
    >>   37 POP_TOP        
    >>   38 LOAD_CONST          3 (None)
         41 RETURN_VALUE   

In other words, in interactive mode, the compiler emits a
PRINT_EXPR bytecode after every expression -- including,
apparently, ones nested within another statement.

> Doesn't it produce loads of debug-like output when pasting
> more complex things?

The fact that nobody seems to have noticed this before
(or at least not complained about it) suggests that people
don't usually paste complex things in interactive mode.

Greg



From meera_bavadekar at hp.com  Wed Nov 10 21:01:40 1999
From: meera_bavadekar at hp.com (Meera)
Date: Wed, 10 Nov 1999 18:01:40 -0800
Subject: How to perform syntax check in Python
Message-ID: <382A2384.CC37FBA8@hp.com>

Hello everybody,

I am looking for  a utility to perform syntax validation in python.
Consider following code -

try:
  a = 3
  print a
  #Do some more processing
except:
  print b

The fact that variable b is not defined is known only when we get
exception.
We will get 'NameError exception' _not_  the actual exception we are
expecting.

When we have multi-developer and multi-platform environment - such
feature is
crucial.  This is one area which I would like python to improve over
conventional language.


Thanks,
Meera




From mal at lemburg.com  Mon Nov  1 09:27:57 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Mon, 01 Nov 1999 15:27:57 +0100
Subject: os.popen() and (IDLE, PythonWin) ?
References: <1270664157-24361801@hypernet.com>
Message-ID: <381DA36D.B04DF056@lemburg.com>

Gordon McMillan wrote:
> 
> Marc-Andre wrote:
> > "M.-A. Lemburg" wrote:
> > >
> > > [os.popen('ver') fails under IDLE and PythonWin, yet works
> > > under "normal"
> > >  Python]
> >
> > Ok, I found the FAQ entry on this:
> >
> > 7.13. Why doesn't os.popen() work in PythonWin on NT?
> [snip]
> > Under PythonWin this is no problem (although I wish os.popen
> > would automatically choose the different approach for me in the
> > above setup). Yet under IDLE I have a problem: win32 doesn't have
> > to be installed... is there some other workaround ?
> 
> tempfile and redirection :-(.

Ok, but how can I prevent the shell windows
from popping up ? I read something about os.spawnv() in the docs,
but the mode argument remains a complete mystery and I always
get IOErrors about files not being found...

Any help appreciated :-),
-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                    60 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/





From asc at bsiag.com  Tue Nov  2 10:49:58 1999
From: asc at bsiag.com (Alex Schroeder)
Date: 02 Nov 1999 16:49:58 +0100
Subject: Download Python Imaging Library Handbook
Message-ID: <uso2o6ggp.fsf@bsiag.com>

Where can I download the Python Imaging Library Handbook so that I can
work with the library without being connected to the net?  I was
unable to find a tarball on www.pythonware.com and www.python.org...

Alex.
-- 
http://www.geocities.com/kensanata/
http://www.bsiag.com/



From vespe at cs.tut.fi  Tue Nov 30 00:41:19 1999
From: vespe at cs.tut.fi (Vespe Savikko)
Date: 30 Nov 1999 07:41:19 +0200
Subject: Htmlizing text
References: <943909426snz@vision25.demon.co.uk>
Message-ID: <vphiu2ktu2o.fsf@hiirihaukka.cs.tut.fi>

Also sprach philh at vision25.demon.co.uk (Phil Hunt):

  Is there a function in the standard Python library to HTML-ize text,
  i.e. to replace 'a > b & c < d' with 'a &gt; b &amp; c &lt; d'?
  
  (I realize this can be done with regular expressions, but is there
  an easy way?)

>>> import cgi
>>> cgi.escape( 'a > b & c < d' )
'a &gt; b &amp; c &lt; d'

  ++Vespe
  
-- 
  ------------------------------------------------------------------
       Vespe Savikko     vespe at cs.tut.fi     - to doom de doomsday -



From cgr at formulae.co.uk  Fri Nov  5 20:02:44 1999
From: cgr at formulae.co.uk (Colm Rafferty)
Date: Sat, 06 Nov 1999 01:02:44 GMT
Subject: smtplib.py how to ?
Message-ID: <38237da0.3456244@news.force9.net>

Dear all,
I cannot insert text into  the headers To: From: CC: etc, of the email
The script below ( the sample in the smtplib.py module) only sends all
the text in the body of the email.

How do I insert text in the headers?

----------------------------------------
import smtplib
s=smtplib.SMTP("localhost")
tolist=["cgr at formulae.co.uk","c.rafferty at virgin.net","cgr at pangea.co.uk"]
msg = '''
From: test at my.org
Subject: testin'...

This is a first test '''

s.sendmail("me at mydomain.co.uk",tolist,msg)
s.quit()

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


Thanks for any help

Colm Rafferty



From ejohnson at netcom.com  Mon Nov 15 22:51:27 1999
From: ejohnson at netcom.com (Eric Johnson)
Date: 16 Nov 1999 03:51:27 GMT
Subject: Bug? - win32pipe.popen2 is moody under W98
Message-ID: <80qkbv$2m5$1@nntp5.atl.mindspring.net>

I'm working on a multi-threaded make engine under Python.  In so
doing, I'm desperately trying to get a popen like facility running 
under Win98 in Python.  I've got Mark Hammond's win32 extensions
(build 127 to be exact), and I'm still striking out.

The following sample code spins off a couple of threads.  Each thread
makes a popen2 call and displays the results.  Ideally, each popen2
call should be capable of being executed in tandem with the rest.


import threading
import thread
import win32pipe

def doit():
  print "here\n";
  a=win32pipe.popen2("dir")
  s=a[1].readline()
  while(s):
    print s
    s=a[1].readline()
  print "Done\n"

for n in range(4):
  a1=threading.Thread(target=doit)
  a1.start()


This and its variant that uses os.popen directly works just fine under
NT 4.0.  But under W98, it gets all hung up on itself quite quickly.
Repeated investigation and experimentation shows that the hang point
appears to be on simulataneous calls to .readline() while another
thread is invoking popen.

Can anyone say either way whether this can be expected to work under
W98?

-Eric Johnson






From mgm at unpkhswm04.bscc.bls.com  Fri Nov 19 09:07:49 1999
From: mgm at unpkhswm04.bscc.bls.com (Mitchell Morris)
Date: 19 Nov 1999 14:07:49 GMT
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <80pe77$osc$1@hub.org> <slrn830he9.n28.mgm@unpkhswm04.bscc.bls.com> <80vp2m$275$1@nntp9.atl.mindspring.net>  <slrn838ohs.fv8.mgm@unpkhswm04.bscc.bls.com> <slrn8393kv.f1h.thantos@chancel.org>
Message-ID: <slrn83amdj.fsi.mgm@unpkhswm04.bscc.bls.com>

In article <slrn8393kv.f1h.thantos at chancel.org>, Alexander Williams wrote:
>On 18 Nov 1999 20:31:58 GMT, Mitchell Morris
><mgm at unpkhswm04.bscc.bls.com> wrote:
>
>>Reading is not so bad ... searching blows. As an example, I'm looking at the
>>page for the "re" module and want to see the argument list for "unpack" but
>>can't seem to remember if it's in "string" or "struct" or
>>Guido-only-knows-where. Lynx provides squat for this operation, so I wind up
>
>Magic word: index.
>
>See the last page of the doc list in HTML, an index of functions and
>terms hyperlinked to the discussion thereof.  Voila.
>
>Has the dimmunation of print media really caused us to forget such
>simple methods?



"For the sake of argument I'll ignore all your fighting words" -- Larry Wall



Your editing of my paragraph removed the portion where I stated that I had to
traverse the document hierarchy. That is, while viewing the "re" page, I must
go "back" to the previous node, in either a browser of the GNU info viewer,
and then forward to another page which contains the information I desire. Is
this not the procedure you're suggesting? Please note that traversing the
hierarchy does not imply that there are no indices nor does it imply that the
indices are unused.

I just find it tedious to have to visit the index page and manually search
(or initiate an automated search) to find the other page of interest. I was
observing that the other "P*" language supplies a tool that removes the need
to visit the index page in between, and I thought that was a good thing.



Or are you saying that your copies of the index pages have all the references
flattened therein so that you can use Lynx's "search in document" feature to
jump directly from the "re.compile" secion to the "struct.unpack" section?



index-pages-are-for-automated-tools-ly y'rs,
+Mitchell

-- 
Mitchell Morris

Any sufficiently complicated C or Fortran program contains an ad-hoc,
informally-specified bug-ridden implementation of half of Common Lisp.
	-- Phil Greenspun



From mlh at vier.idi.ntnu.no  Mon Nov  1 10:55:02 1999
From: mlh at vier.idi.ntnu.no (Magnus L. Hetland)
Date: 01 Nov 1999 16:55:02 +0100
Subject: How can I make my assertions smarter?
References: <381D7087.BF8510E5@webone.com.au>
Message-ID: <y0jpuxuryuh.fsf@vier.idi.ntnu.no>

Stuart Hungerford <stuart.hungerford at webone.com.au> writes:

> Hi all,
> 
> I'm using some simple functions to implement crude precondition,
> postcondition and invariant checking in Python classes.
> 
> (Tim: please be gentle ;-)
> 
> Is there some neat way to wrap these tests up so I can print any
> code expression that fails, e.g:
> 
>      Precondition failed foo.py line 42: spoon.shape == 'bent'
> 
> which would appear at line 42 of foo.py as
> 
>      pre(spoon.shape == 'bent')

How about using (assuming that you won't use assert):

       pre("spoon.shape == 'bent'")

And then using eval() in your code? Like...

  def pre(condition):
      if not eval(condition):
          raise "Precondition Failure", condition

Pretty simple, IMO...

--

  Magnus          Echelon jamming noise:
  Lie             FBI CIA NSA Handgun Assault Bomb Drug Terrorism
  Hetland         Special Forces Delta Force AK47 Hillary Clinton 



From a.eyre at optichrome.com  Fri Nov 26 04:30:51 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Fri, 26 Nov 1999 09:30:51 -0000
Subject: XML to in-memory tree?
In-Reply-To: <tRf%3.8219$cz3.293110@nnrp1.uunet.ca>
Message-ID: <000001bf37f0$edcc4850$3acbd9c2@peridot.optichrome.com>

> Hi.  Has anyone built a layer on top of XMLLIB that will load an XML
> document into an in-memory (similar to the one built by explicit calls to
> HTMLgen)?

Have a look through the XML-sig's code. There's a SAX DOM builder if that's
what you mean.

See: http://www.python.org/topics/xml/ or
http://www.python.org/sigs/xml-sig/

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com
--------------------------------------------




From davidh at progmatics.com.au  Thu Nov 25 18:59:20 1999
From: davidh at progmatics.com.au (David Hobley)
Date: Fri, 26 Nov 1999 10:59:20 +1100
Subject: oddity with ftplib
Message-ID: <383DCD58.44915F58@progmatics.com.au>

Has anyone had any problems using the storbinary command of
ftplib?

I had been using it for a while for transferring image files to
an ISP. I recently changed ISP and the new ISP appears to only
get truncated images when it is above a certain size.

I am looking into the problem, but thought i would check if
anyone had seen this before?

-- 
Cheers,
david                                 davidh at progmatics.com.au

Progmatics Pty Ltd - Architects of IT and Internet Solutions

Level 8, 191 Clarence Street              Phone +61 2 9262 4933
Sydney NSW Australia                      Fax   +61 2 9262 4045
http://www.progmatics.com.au/



From phd at phd.russ.ru  Thu Nov 25 07:01:47 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Thu, 25 Nov 1999 12:01:47 +0000 (GMT)
Subject: Python complaints
In-Reply-To: <383CC835.B585EBDD@Lugoj.Com>
Message-ID: <Pine.LNX.4.21.9911251159140.11528-100000@fep132.fep.ru>

On Wed, 24 Nov 1999, James Logajan wrote:
> def ifexp(cond, trueVal, falseVal):
>    if cond:
>       return trueVal
>    else:
>       return falseVal
> 
> print "You scored %s point%s" % (score, ifexp(score == 1, "", "s"))

   The problem with this solution is that python will calculate both
expressions (trueVal, falseVal) even if "cond" is false - what is VERY
undesireable in most situations. What if I just cannot calculate falseVal
if "cond" is false?

Oleg.
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From spaceram00 at earthlink.net  Tue Nov  9 03:35:50 1999
From: spaceram00 at earthlink.net (Spaceram)
Date: Tue, 9 Nov 1999 00:35:50 -0800
Subject: Grail Brower question
Message-ID: <808m9h$nnn$1@holly.prod.itd.earthlink.net>

For anyone who knows what the grail browser is I have a question.
How do I use the brower? I download it and when i doubleclick it it opens a
open with box which means that windows doesn't know what to do with the
file.

Please email all answers to spaceram00 at earthlink.net






From peter at hotmail.com  Fri Nov  5 16:17:47 1999
From: peter at hotmail.com (peter at hotmail.com)
Date: Friday, 05 Nov 1999 15:17:47 -0600
Subject: No banner,free,homepage storage with high speed 15608
Message-ID: <05119915.1747@hotmail.com>

No banner,free,homepage storage with high speed. home.ublaze.com

http://home.ublaze.com




From nascheme at enme.ucalgary.ca  Mon Nov  8 14:57:24 1999
From: nascheme at enme.ucalgary.ca (Neil Schemenauer)
Date: Mon, 08 Nov 1999 19:57:24 GMT
Subject: Fastest GUI toolkit?
References: <01bf29f1$cbfd5e60$5a41c3d1@Cyberus.cyberus.ca>
Message-ID: <slrn82eapd.pct.nascheme@adler.dynodns.net>

Gordon Williams <g_will at cyberus.ca> wrote:
>
>I have an application where I need to draw many lines fast and repeatedly
>(an oscilloscope).  The number of lines ranges from about 1000 to 16000. 
>Currently I have the application using Tkinter but I am finding it too
>slow.

I assume you are using the canvas to draw.  I think it is much
faster to use Canvas.coords to move the lines then to redraw
them.  If this approch works for your application it may be worth
a try.


    Neil



From fredrik at pythonware.com  Mon Nov 22 09:37:33 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 22 Nov 1999 15:37:33 +0100
Subject: RDF/RSS
References: <943225647snz@vision25.demon.co.uk>
Message-ID: <00e301bf34f7$1db60da0$f29b12c2@secret.pythonware.com>

Phil Hunt <philh at vision25.demon.co.uk> wrote:
> RDF channels

you mean RSS channels, right?

    http://my.netscape.com/publish/help/quickstart.html

(or are you talking about something completely
different? -- the reason I'm asking is that some
moron just flamed me because a Python solution
I posted to this newsgroup didn't work in C++...).

> Oh, and what does RDF stand for anyway?

resource description framework:

    http://www.w3.org/RDF/

(iirc, the last version of RSS is no longer RDF-
compatible, but that's another story).

</F>




From quinn at riyal.ugcs.caltech.edu  Sat Nov 27 21:22:57 1999
From: quinn at riyal.ugcs.caltech.edu (Quinn Dunkan)
Date: 28 Nov 1999 02:22:57 GMT
Subject: Binary vs non-binary
References: <19991127103504.A1536@stopcontact.palga.uucp>
Message-ID: <slrn8414hl.eqn.quinn@riyal.ugcs.caltech.edu>

On Sat, 27 Nov 1999 10:35:04 +0100, Gerrit Holl <gerrit.holl at pobox.com> wrote:
>Hello,
>
>can someone explain me the difference in opening a file in binary mode
>and opening it not in binary mode? And when is some file defined to be
>binary, when it contains an ASCII 0? If so, I understand you need to say
>it's binary in C, but if Python doesn't store it strings with an ASCII 0
>on the end, why is there any difference?

This is an artifact of C having grown up on unix, where lines end with \n,
and then being used on other systems, where lines end with \r\n (also, many
network protocols use \r\n for some reason).  Anyway, if you open a file
in C on, say, DOS, there is going to be an inconsistency because C wants
to think of lines being terminated by single characters.  So stdio does some
behind the scenes munging to turn all \r\ns into \ns and vice-versa.  But if
you did that on binary data, you'd corrupt it because \n is just another bit
pattern there.  So you need to know whether you're opening a binary file or
not.  And since python is written in C..

Basically: under unix and other systems that use \n, you don't have to worry
about binary mode---setting it does nothing.  But if you want your programs to
be portable to DOS (and probably MacOS too, which uses \r), you should use
binary mode when opening binaries.

It has nothing to do with ASCII 0 (NUL), although it's a good thing to
notice: many C programs choke on input with NULs in it.

*whew*

remember-that-when-breaking-into-the-pentagon-ly y'rs



From htrd90 at zepler.org  Fri Nov  5 14:54:34 1999
From: htrd90 at zepler.org (Toby Dickenson)
Date: Fri, 05 Nov 1999 19:54:34 +0000
Subject: A Comparison of Python and Ruby Extension Interfaces (Re: Comparison   between Python and "Ruby")
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp> <7vjtrc$hmm$1@newshost.accu.uu.nl> <38204614.2EAC6051@compaq.com> <021119992327527678%pereira@research.att.com> <38215CE7.70C3D560@compaq.com>
Message-ID: <SzUjOCp3Jbh1lZd9WdVDdXeTiSia@4ax.com>

Greg Ewing <greg.ewing at compaq.com> wrote:

>Fernando Pereira wrote:
>>
>> Same comment. There is no portable way of unwinding the C stack. What
>> does that mean for Ruby's portability?
>
>It uses setjmp/longjmp. Presumably these work correctly on
>any system where they are implemented. Porting Ruby to a
>system where they're not implemented could be difficult.
>

Presumably this makes it difficult for Ruby extensions to use C++?


Toby Dickenson



From morse at harborcom.net  Tue Nov  9 08:01:50 1999
From: morse at harborcom.net (Kevin Dahlhausen)
Date: Tue, 09 Nov 1999 13:01:50 GMT
Subject: Windows Sound Card
References: <01bf2a01$27e7cda0$8355c3d1@Cyberus.cyberus.ca>
Message-ID: <38281a8e.88096035@news.oh.verio.com>

I'd suggest finding a C/C++ library on the net or a toolkit header file from the
vendor.  Then take a look at SWIG.  SWIG will generate a python module from the
C/C++ header files, sometimes without any extra work.  O yeah, then make the
module available somewhere so Python can have a voice on lots of PC's!
(What  kind of sound does a Python make anyway?)

>I am looking for information on a module to get data from the input from a
>sound card on a Windows box.  Is this possible using Python??  I have had a
>search through all the docs and FAQs but havent found anything for Windows.
>
>Any pointers would be a help.
>
>Thanks,
>
>Gordon Williams




From charliehubbard76 at yahoo.com  Mon Nov 15 15:57:41 1999
From: charliehubbard76 at yahoo.com (charliehubbard76 at yahoo.com)
Date: Mon, 15 Nov 1999 20:57:41 GMT
Subject: Python, COM and Excel
Message-ID: <80ps42$kkk$1@nnrp1.deja.com>

Hello,

I'm trying to write a python script to take an excel
workbook and turn it into html pages.  I'm trying to
automate this process, but I can't seem to get the
bounds of the selection through com.  I can get the
contents of a selection with

app = Dispatch(Excel.Application)
app.Selection()

But this just gives me the data contained with the cells.
I want to be able to check fonts, bold, and format
information of each cell to get it as close to the excel
formatting as possible.  In order to do this I need to
either get the bounds of the selection, or retrieve all
the cells references' being used.  Preferrably a bounding
selection of all the cells being used.  For example,
"A3:AN32" or "1,3,36,32".  Any ideas?

thanks
charlie


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



From aahz at netcom.com  Wed Nov 24 11:05:44 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 24 Nov 1999 16:05:44 GMT
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <8666ysfta9.fsf@ix.netcom.com>
Message-ID: <81h2co$96s$1@nntp9.atl.mindspring.net>

In article <8666ysfta9.fsf at ix.netcom.com>,
Brad Knotwell  <knotwell at ix.netcom.com> wrote:
>
>In general, the only thing I've found annoying about Python are the scoping
>rules:

That's actually one of the things I like best about Python.
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)

Plus ca change, plus c'est la meme chose.



From fdrake at acm.org  Wed Nov 10 16:18:08 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Wed, 10 Nov 1999 16:18:08 -0500 (EST)
Subject: random function
In-Reply-To: <3829D78B.7FC142F4@scripps.edu>
References: <3829D78B.7FC142F4@scripps.edu>
Message-ID: <14377.57616.255562.594683@weyr.cnri.reston.va.us>

Olivier Roche writes:
 > By looking at the index of the python documentation index i haven't
 > seen any random function description. Could somebody tell me if a
 > random function to generate random number (between 0 and 1) exists

  I see the following entries in the latest version; take a look at
the docs for both the random and whrandom modules:

        random() (in whrandom)
        random (standard module)


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From fdrake at acm.org  Fri Nov  5 15:13:19 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Fri, 5 Nov 1999 15:13:19 -0500 (EST)
Subject: (no subject)
Message-ID: <14371.14943.544121.316575@weyr.cnri.reston.va.us>

   
  IPC8 Call For Demos and Posters
  
  Important Dates
  
     * Nov. 30 : Deadline for demos and posters.
     * Dec. 15 : Notification of acceptance for demos and posters.
       
  Demonstrations and Posters
  
   We invite researchers and developers to present their work in
   demonstrations or posters. Works in progress and results which don't
   warrant a full paper may be presented this way, as well as
   demonstrations of interesting applications.
   
   Demos and posters will be selected based on the broad applicability of
   new technologies and novelty of new approaches.
   
   Tables and power will be provided. Computers will not be provided; if
   you intend to do a demonstration, bring your own with all the software
   you need. There may be a slow Internet connection (at most a shared
   ISDN line); please let us know if Internet access is important to your
   demonstration.
   
  What is a poster?
  
   When presenting a poster, you typically bring one or more posters that
   can be tacked up to a wall. The posters should provide a concise
   summary of your work, and diagrams or other graphics are strongly
   encouraged. You then stand next to your poster and discuss your work
   with other conference attendees (who will undoubtedly ask why you're
   standing there).
   
   If you're feeling confident, you can provide handouts that explain
   your research or project in more detail, or tell where to find more
   information. Bring plenty of copies.
   
  So, what's a demo?
  
   A demo is a poster, but you use your computer to actually show people
   what you did. You can have a slide presentation or let people play
   with your applications.
   
   If you want to give a demonstration, be sure to bring your computer!
   
  What to Submit and How to Submit It
  
   To request space for a demonstration or poster, send an email to Fred
   Drake (fdrake at acm.org) describing your demonstration or poster. If you
   have special space or power requirements, please be sure to include
   the appropriate information. If you require Internet connectivity, be
   sure to tell us about that as well!
   
  For more information
  
   Questions regarding demo and poster submissions should be addressed to
   Fred Drake (fdrake at acm.org).



From philh at vision25.demon.co.uk  Mon Nov 22 15:01:09 1999
From: philh at vision25.demon.co.uk (Phil Hunt)
Date: Mon, 22 Nov 99 20:01:09 GMT
Subject: RDF/RSS
References: <943225647snz@vision25.demon.co.uk> <00e301bf34f7$1db60da0$f29b12c2@secret.pythonware.com>
Message-ID: <943300869snz@vision25.demon.co.uk>

In article <00e301bf34f7$1db60da0$f29b12c2 at secret.pythonware.com>
           fredrik at pythonware.com "Fredrik Lundh" writes:
> Phil Hunt <philh at vision25.demon.co.uk> wrote:
> > RDF channels
> 
> you mean RSS channels, right?
> 
>     http://my.netscape.com/publish/help/quickstart.html

That's it.

> (or are you talking about something completely
> different? -- the reason I'm asking is that some
> moron just flamed me because a Python solution
> I posted to this newsgroup didn't work in C++...).
> 
> > Oh, and what does RDF stand for anyway?
> 
> resource description framework:
> 
>     http://www.w3.org/RDF/
> 
> (iirc, the last version of RSS is no longer RDF-
> compatible, but that's another story).

What is the latest version number? I've heard of 0.9 and 0.91,
but my.netscape.com only seems to describe 0.9.

-- 
Phil Hunt - - - phil at comuno.com
Want to see the FAQs for your favourite newsgroups? Read them on
the web with Comuno's FAQ server at http://www.comuno.com/faq/ 




From rmeegan at murl.com  Mon Nov  8 11:29:01 1999
From: rmeegan at murl.com (Robert Meegan)
Date: Mon, 8 Nov 1999 10:29:01 -0600 (EST)
Subject: pythonistas or pythoneers?
In-Reply-To: <07111999.2@sanctum.jae>
Message-ID: <Pine.LNX.4.10.9911081026470.4450-100000@robert.polytopic.com>

On Sun, 7 Nov 1999, Juergen A. Erhard wrote:

> 
> I'm confused: are we pythoneers or pythonistas?
> 
> I tend to go for pythoneers... sounds more pythonic.
> 
> Or are these two levels of experience?  If so, what's higher?
> 
> Confused,
> 

I always thought we were Pythogorians.


Your most righteous servant,

--- Robert


Robert Meegan

Murl.com - 
Because it's not just your data, it's your life.




From I.Clarke at strs.co.uk  Thu Nov  4 06:25:24 1999
From: I.Clarke at strs.co.uk (Ian Clarke)
Date: Thu, 04 Nov 1999 11:25:24 +0000
Subject: A morality tale of Perl versus Python
References: <199911041120.LAA22431@tardis.tardis.ed.ac.uk>
Message-ID: <38216D24.89E39380@strs.co.uk>

>From rec.humor.funny:

> ___________________________________________________________________________
> EXTERIOR: DAGOBAH -- DAY
>            With Yoda strapped to his back, Luke climbs up one of the
>         many thick vines that grow in the swamp until he reaches the
>         Dagobah statistics lab. Panting heavily, he continues his
>         exercises -- grepping, installing new packages, logging in as
>         root, and writing replacements for two-year-old shell scripts
>         in Python.
> 
> YODA: Code!  Yes.  A programmer's strength flows from code maintainability.
>       But beware of Perl.  Terse syntax... more than one way to do it...
>       default variables.  The dark side of code maintainability are they.
>       Easily they flow, quick to join you when code you write.  If once
>       you start down the dark path, forever will it dominate your destiny,
>       consume you it will.
> 
> LUKE: Is Perl better than Python?
> 
> YODA: No... no... no.  Quicker, easier, more seductive.
> 
> LUKE: But how will I know why Python is better than Perl?
> 
> YODA: You will know.  When your code you try to read six months from
>       now.
> ___________________________________________________________________________
> 
> --
> Selected by Jim Griffith.  MAIL your joke to funny at netfunny.com.
> Attribute the joke's source if at all possible.  A Daemon will auto-reply.
> If you don't need an auto-reply, submit to rhf at netfunny.com instead.
> For the full submission guidelines, see http://www.netfunny.com/rhf/
> 
> This joke's link: http://www.netfunny.com/rhf/jokes/99/Nov/perl.html

--
Ian Clarke                        Email: I.Clarke at strs.co.uk
Homepage: http://www.gnu.demon.co.uk/
Also see: http://freenet.on.openprojects.net/
"All we see and all we seem is but a dream within a dream"



From jeremy at cnri.reston.va.us  Fri Nov  5 15:02:49 1999
From: jeremy at cnri.reston.va.us (Jeremy Hylton)
Date: Fri, 5 Nov 1999 15:02:49 -0500 (EST)
Subject: IPC8: SouthWest Airlines Internet specials
In-Reply-To: <Pine.WNT.4.04.9911051036270.260-100000@rigoletto.ski.org>
References: <Pine.WNT.4.04.9911051036270.260-100000@rigoletto.ski.org>
Message-ID: <14371.14313.700685.807764@goon.cnri.reston.va.us>

Everyone planning to fly to the conference should consider flying
through BWI.  I have flown from BWI to San Jose and San Francisco a
few times this year, each time for about $200.  Flights through BWI
are often cheaper than flights from Dulles or National.

Jeremy




From meh9 at cornell.edu  Tue Nov  2 10:18:59 1999
From: meh9 at cornell.edu (Matthew Hirsch)
Date: Tue, 02 Nov 1999 10:18:59 -0500
Subject: Printing strings
Message-ID: <381F00E1.721BB47F@cornell.edu>

Hi All,

I'm trying to do something like the following:

a=4.0
b=2.0
print '2(%3.1f)' % a,b

where I really mean

print '%3.1f %3.1f' % a,b

but it's printing

2(4.0) 2.0

Is there a way of doing this so I don't have to write out %3.1f,
multiple times?

Thanks,
Matt





From clarence at beach.silcom.com  Sat Nov  6 20:27:59 1999
From: clarence at beach.silcom.com (Clarence Gardner)
Date: Sun, 07 Nov 1999 01:27:59 GMT
Subject: A Modest Proposal -- <wink>
References: <99110613580407.02970@quadra.teleo.net>
Message-ID: <zA4V3.31$An3.1679@newsfeed.avtel.net>

Patrick Phalen <python-list at teleo.net> wrote:
: In XHTML, this has discrepancy been ameliorated by adding the closing
: delimiter "/>" to "<br>", to yield "<br/>". This indicates to the
: XML processor that the tag is EMPTY. Therefore, under the new proposal,
: "<wink>" would, of necessity, become "<wink/>".

In fact, since <BR>, while empty, does admit of attributes, XHTML must
allow tokens to follow the slash, and this allows the very important
fractional winkery to be spelt also.  I'm sure we can get the TimBot to
change to using divisors rather than multipliers. <wink/10>

-- 
Clarence Gardner
Software Engineer
NetLojix Communications    NASDAQ:NETX
clarence at netlojix.com



From python-list at teleo.net  Tue Nov  9 21:14:36 1999
From: python-list at teleo.net (Patrick Phalen)
Date: Tue, 9 Nov 1999 18:14:36 -0800
Subject: Java vs. Python - Serialization Bakeoff
Message-ID: <99110918085909.01567@quadra.teleo.net>

http://starship.python.net/crew/gandalf/DNET/serial/jp-serial.html
contains an report on tests comparing Java's and Python's data
marshalling, using various criteria. Numerous interesting charts and
graphs are presented.

Since this was published in 1996 (Python 1.4), I wonder if anyone
expert in this subject has any comment on whether the intervening years
would have changed the conclusions substantively -- namely that Python
is, on average, faster, simpler, and more flexible.



From robin at jessikat.demon.co.uk  Tue Nov  9 05:31:11 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Tue, 9 Nov 1999 10:31:11 +0000
Subject: python: something completely different
References: <3820C0B8.62B2@iowegian.com> <38249ED2.BB7EA1DE@znet.com>
 <8028jv$7cf$1@hub.org> <38259F18.437@iowegian.com>
 <38278B5F.1AFF953C@python.net> <3827514F.3A8F@iowegian.com>
Message-ID: <f6VqjEAvf$J4EwAl@jessikat.demon.co.uk>

In article <3827514F.3A8F at iowegian.com>, Grant Griffin
<grant.griffin at iowegian.com> writes
....
>One could take a cue from Octave here.  Octave provides Matlab-style
>plot commands which are simply wrappers to Gnuplot.  (However, Octave's
>reliance on Gnuplot creates a significant installation barrier in the
>Windows world.)
>
>Speaking as someone who is very new to Python, I am impressed by the
>variety of plotting schemes available for it, as well as somewhat
>bewildered by them.  From a new user's point of view, one would like to
>be handed "standards" (while still having choices).  Python's use of Tk
>is a good example of this balance relative to GUI.  I believe that the
>inclusion of some sort of high-quality, portable, easy-to-install,
>standardized plotting system would go a long way to helping Python
>invade Matlab's turf.  Perhaps one of the plotting systems currently
>available meet all these needs, if only it were given the blessing by
>being distributed with NumPy.
>
>> 
....
There are a lot of these available it's true, but few of them will print
well from win32. Fewer still will do decent 3d charts of any kind. If I
could just get to the ease of use of the graph tools embedded inside M$
tools I would be happy.
-- 
Robin Becker



From jon_mclin at attglobal.net  Tue Nov 30 07:18:02 1999
From: jon_mclin at attglobal.net (Jon McLin)
Date: Tue, 30 Nov 1999 12:18:02 +0000
Subject: Installation Problem (1.5.2)
References: <3843B816.F4A31F07@attglobal.net> <m366ykryi1.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <3843C07A.21A33E57@attglobal.net>

Ahhh.  Much better now.  Thanks.

Incidentally, I did in fact type "exec_prefix", which was accepted (i.e., the
binary was placed in /usr/bin).  I'm not going to go find out why just now...

But simply setting --prefix to /usr solved my problem.  Must be the lack of
caffeine this morning.

Thanks again,
Jon


Michael Hudson wrote:

...

> 1) you mean --exec-prefix.
>
> 2) you may have binaries in /usr/bin, but you also have libraries in
>    /usr/lib, not /usr/local/lib. pass --prefix=/usr to configure.
>
> Regards,
> Michael




From mlh at vier.idi.ntnu.no  Sat Nov 13 11:10:42 1999
From: mlh at vier.idi.ntnu.no (Magnus L. Hetland)
Date: 13 Nov 1999 17:10:42 +0100
Subject: A Modest Proposal -- <wink>
References: <99110613580407.02970@quadra.teleo.net> <8072o6$m4i$1@newshost.accu.uu.nl> <y0jiu3b2357.fsf@vier.idi.ntnu.no> <80jqqo$n3m$1@newshost.accu.uu.nl>
Message-ID: <y0j9042tlrx.fsf@vier.idi.ntnu.no>

m.faassen at vet.uu.nl (Martijn Faassen) writes:

> Magnus L. Hetland <mlh at vier.idi.ntnu.no> wrote:
> > m.faassen at vet.uu.nl (Martijn Faassen) writes:
> 
> >> 
> >> As you can see, XWTSNTSDL is fully consistent with your current proposals. 
> >> It is therefore expected the major internet browsers will soon be able to
> >> support these new standards.
> 
> > It is not consistent with XML... It would have to be <wink/>....
> 
> Ah, but XWTSNTSDL takes care of that, of course. :)
> 
> It defines a XPPMS (eXtensible PreProcessor Macro System), which allows
> single <wink> tags. The XPPMS system transforms the source into valid XML.
> After that the other subsystems of XWTSNTSDL take over.

But - if you have XPPMS, why do you need XWTSNTSDL? You could just
transform

   This is some text <wink>

into

   <wink>This is some text</wink>

and Bob's your uncle... Then you can use standard XML tools... :)

Or - does XWTSNTSDL include tools for deciding where the wink-section
starts, semantically?

> I-can-come-up-with-more-acronyms-than-you-can-with-objections-ly yours, 
> 
> Martijn
> 

--

  Magnus          Echelon jamming noise:
  Lie             FBI CIA NSA Handgun Assault Bomb Drug Terrorism
  Hetland         Special Forces Delta Force AK47 Hillary Clinton 



From c.hintze at gmx.net  Wed Nov 17 03:13:36 1999
From: c.hintze at gmx.net (Clemens Hintze)
Date: Wed, 17 Nov 1999 08:13:36 GMT
Subject: Type/Class Unification in Ruby
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp> <7vjtrc$hmm$1@newshost.accu.uu.nl> <38204614.2EAC6051@compaq.com> <5ld7trtak5.fsf@eric.cnri.reston.va.us> <3826D63C.2731BB1F@compaq.com> <38328948.9E35556F@compaq.com>
Message-ID: <m3d7t9v8lr.fsf@qiao.localnet>

Hi,

Greg, you have done a great analyze! I wish you would have done it
sooner ;-) It would have spare me some time to fiddle that out on my
own. This part of Ruby is, unfortunately, not too well documented. :-(

But in addition to your analyze, may I add some remarks?

>>>>> "Greg" == Greg Ewing <greg.ewing at compaq.com> writes:

[...]

    Greg> 2) A class can only inherit from at most one other class.

This is, in fact, true. But it is worth mentioning, that Ruby features 
single inheritance *by*purpose*. Not because of the implementation
details!

If it would implement multi-inheritance, the class structure could
e.g. use an array of pointer to all such C structs, not only one
pointer, as implemented now!

    Greg> The only methods which actually create C structures are the
    Greg> built-in instantiation methods of the built-in classes, so
    Greg> even if you define your own instantiation method in your
    Greg> subclass, it eventually has to call a built-in one somewhere
    Greg> up its inheritance chain, which knows what kind of C
    Greg> structure is needed.

That is not totally true (or, perhaps, I misunderstood you). You have
to differ two kind of class-building. you seem to describe, how Ruby's 
built-in classes are implemented. But there is also a second way, that 
should be used by own extensions. *Only* Ruby built-in (means built
into the interpreter) classes, should use the way you have described.

The second (preferred) way is, to build instances of class DATA
wrapping the C struct you have allocated for your extension. That DATA
instance is marked as beeing not a DATA instance, but an instance of
your extension class! So that means, that *you* would allocate your
instance!

Deriving a Ruby class coded in C from another one also coded in C,
seems to be the only right tricky part of that language, IMHO. :-)

    Greg> This scheme will fail if you somehow manage to extract a
    Greg> built-in method out of one class and stuff it into
    Greg> another. However there doesn't seem to be any obvious way to
    Greg> do this (nothing like myClass.__dict__["foo"] in Python). So
    Greg> it seems to be safe.

Good derivation! I have also missed it for some time. But as Ruby
features closures, I had learned to use them instead of extraced
methods. Anyway Ruby has no functions, only methods.

You could extract a method object (means a method implementation). But
you would not be able to use it out-of-context of the class it was
taken from! As matz (the author of Ruby) has told me, methods are too
strong connected to the class they belongs to (performance reasons).

    Greg> Now, what happens if you assign to a new instance variable
    Greg> in a method of your subclass of a built-in class? At first
    Greg> sight it seems that this can't possibly work, because the C
    Greg> structs of built-in types don't have any place to store
    Greg> arbitrary instance variables.

;-)))))))

    Greg> When I tried it, however, to my surprise, it worked!  It
    Greg> turns out that there is a global table whose purpose is to
    Greg> hold the values of these instance variables (in Python terms
    Greg> it is a dictionary of dictionaries, indexed by
    Greg> [object][variable_name]). Python might be able to use a

That was a great idea from matz, IMHO! :-)

    Greg> similar solution, although it might cause ref counting
    Greg> problems.

Once again, thanks for Ruby's true garbage collection! :-)

    Greg> As a final note, some built-in classes, such as Fixnum,
    Greg> don't have any instantiation method, so although you can
    Greg> subclass them, there's no point because you can never create
    Greg> an instance of your subclass.

In addition to your analyze: in the past, at least the classes of
Numeric and childs had instanstiation methods. But some user of Ruby
have convinced matz, that it would make no sense to derive another
numeric type from them. So the methods were delete. The big exception,
AFAIK, was class Fixnum. It never had any instanciation method, as it
doesn't wrap any struct. The pointer-to-struct is 'misused' as
contents of Fixnum. If I remember right, Smalltalk has also chosen
that way for implementing SmallInteger.

    Greg> Others, however, such as Array (Ruby's version of ListType)
    Greg> do have an instantiation method, and subclasses work just
    Greg> the way you'd expect.

It is worth mentioning too, that it is often not necessary to derive
something from e.g. Array. Often it is better to use the module
mechanism to add methods to class Array or some of its instances
during runtime.

You can do that also in your extension. So if Ruby's Array would not have
'push' and 'pop' methods, I would not derive a class Stack from Array, 
but I would implement a module Stack (implementing 'push' and 'pop')
and simply add them to class Array, whenever I need them ... :-)))

    Greg> Hoping this has been enlightening, Greg

We would need more of such ... :-)

Clemens.



From python-list at teleo.net  Sat Nov  6 14:24:22 1999
From: python-list at teleo.net (Patrick Phalen)
Date: Sat, 6 Nov 1999 11:24:22 -0800
Subject: making the Python case to my dept.
In-Reply-To: <3824723A.5DE35FE5@callware.com>
References: <9911051740130D.02530@william.jinsa.org> <801p3q$q01$1@nntp2.atl.mindspring.net> <3824723A.5DE35FE5@callware.com>
Message-ID: <99110611340304.02970@quadra.teleo.net>

[Ivan Van Laningham, on Sat, 06 Nov 1999]:

:: Aahz Maruch wrote:
:: > I believe that there are more jobs for good Python programmers than
:: > people with the skills to fill them.  My company is certainly having
:: > difficulties (jobs in Mountain View! ask me!), and having HP Cupertino
:: > advertising for 10-12 Python positions only increases the difficulty.

:: I really hope that Aahz is correct.  As most of you know, I'm stuck out
:: here in Utah, which is widely known to be significantly behind *any*
:: curve you want to name.
:: 
:: If there's a curve coming to Utah, I sure hope it will be one with jobs
:: for Pythonistas, because there certainly aren't *any* here now.

Look at it this way: the HP jobs all appear to center around developing
the E-speak technology. Python has been chosen as the official
scripting language for E-speak. Novell has publicly endorsed E-speak. If
Novell decides to builds Netware hooks into E-speak,  they'll probably
hire rafts of Pythonistas in Utah.  :)



From iam at not.you  Wed Nov 17 09:23:51 1999
From: iam at not.you (Ken Power)
Date: Wed, 17 Nov 1999 14:23:51 GMT
Subject: Alphabetized dictionary listings
Message-ID: <3832b6ce.80769668@news1.mysolution.com>

Why does this occur? More to the point, how can I have the print-out
be in alphabetical order?

>>> myList = ['Abel', 'Dexter', 'Francis', 'Matthew', 'Samantha', 'Tyrone']
>>> myDict = {}
>>> for name in myList:
	myDict[name] = 0
	
	
>>> for key in myDict.keys():
	print key
	
	
Tyrone
Samantha
Francis
Abel
Matthew
Dexter


Again, how can I have the printout be in Alphabetical order? I also
noticed that if I fed values to myDict[key] from another list, that
was constructed so it's values would match the keys in myDict, that
the values would be assigned to the wrong keys:

>>> myList2 = ['lebA', 'rexteD', 'sicnarF', 'wehttaM', 'ahtnamaS', 'enoryT']
>>> var = 0
>>> for key in myDict.keys():
	myDict[key] = myList2[var]
	var = var + 1

>>> for key in myDict.keys():
	print key, myDict[key]
	
	
Matthew ahtnamaS
Francis sicnarF
Abel wehttaM
Samantha rexteD
Dexter enoryT
Tyrone lebA
As you can see the assignements are in correct.  Well, if this is
understandable, could someone offer some insight? Thank you.

Ken
--------------------------------
Ken Power
uncle_wiggly at bigfoot dot com
get that?
--------------------------------



From mlh at vier.idi.ntnu.no  Tue Nov 16 05:06:15 1999
From: mlh at vier.idi.ntnu.no (Magnus L. Hetland)
Date: 16 Nov 1999 11:06:15 +0100
Subject: list.without()?
References: <NDBBIAPHEKGOBIHCBLIICEOMCAAA.mcfletch@vrtelecom.com> <y0jiu32iyi8.fsf@vier.idi.ntnu.no> <01b001bf3019$48c5a010$f29b12c2@secret.pythonware.com>
Message-ID: <y0jg0y6iwdk.fsf@vier.idi.ntnu.no>

"Fredrik Lundh" <fredrik at pythonware.com> writes:

> Magnus L. Hetland <mlh at vier.idi.ntnu.no> wrote:
> > Why on *earth* would you use a while/try/break-combination?
> 
> maybe because the meaning of "without a
> specified element" is unclear.  mike's version
> guarantees that:
> 
>     list.without(element)
>     assert element not in list
> 
> never fails, while your version behaves just
> like "remove" -- if there are several instances
> of an element value, only the first one is
> removed:
> 
>     count = list.count(element)
>     list.without(element)
>     assert count == 0 or list.count(element) == count-1

Oh - sorry about that. :)

My intended meaning was that implemented in my own routine, which made
me blind to the benefits of Mike's version. ;)

> 
> in either case, adding "without" (like remove, but
> returns a new list) strikes me as being even more
> contrived than earlier proposals like "sorted" (like
> sort, but returns a new list), etc.

Okay... I don't think that is contrived either, but I guess we just
disagree ;)

> 
> </F>
> 

--

  Magnus          Echelon jamming noise:
  Lie             FBI CIA NSA Handgun Assault Bomb Drug Terrorism
  Hetland         Special Forces Delta Force AK47 Hillary Clinton 



From phd at phd.russ.ru  Tue Nov 16 12:17:42 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Tue, 16 Nov 1999 17:17:42 +0000 (GMT)
Subject: function override???
In-Reply-To: <80s0sd$1g4k$1@rtpnews.raleigh.ibm.com>
Message-ID: <Pine.LNX.4.20.9911161715280.2604-100000@fep132.fep.ru>

On Tue, 16 Nov 1999, Arinte wrote:
> I am defining a python callback for my embedded app, but I want the script
> writer to override my function.  So, basically what I want is a function
> prototype.  Ex. here is what I am thinking.
> 
> ;---fprot.py
> def overrideme(self, str, someint)
> 
> ;--fimp.py
> import fprot   #I know this ain't right.
>                           # I don't know how to import from other python
> scripts yet.  Enlighten me?

   I ithink it is correct. Why do you think it is not?

> def overrideme(self, str, someint)
>     do_some_junk()

   Define your procedure:

;---fprot.py
def overrideme(self, str, someint):
   pass

;--fimp.py
import fprot

   And assign your override function:

fprot.overrideme = my_override

   After this, call fprot.overrideme from any place (C or Python).

Oleg.
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From mwh21 at cam.ac.uk  Sat Nov  6 05:46:12 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 06 Nov 1999 10:46:12 +0000
Subject: exec / execfile namespace problem
References: <004aa0e3.c3e96526@usw-ex0101-002.remarq.com>
Message-ID: <m3904baoej.fsf@atrus.jesus.cam.ac.uk>

Karan Vasudeva <karanvasudevaNOkaSPAM at usa.net.invalid> writes:

> Hi,
> 
> I'm trying to do this:
> execfile('cool.py')
> 
> and this:
> 
> import fileinput
> chunk = ''
> for line in fileinput.input('cool.py'):
>    chunk = chunk + line
> exec(chunk)

How does that differ from 

exec open('cool.py').read()

? What you have will be really quite inefficient, because repeatedly
executing

    chunk = chunk + line

is a quadratic algorithm ... but back to the point.

> 
> where 'cool.py' is this: (in part)
> import httplib
> import time
> import whrandom
> h = httplib.HTTP('')
> h.putrequest('GET', '/favicon.ico')

Have you tried executing `python cool.py'? The obvious (to me) problem
with the above is that you're not telling httplib about server! I
think you want

h = httplib.HTTP("server.domain.tld")

HTH,
Michael



From maxco at telenia.it  Thu Nov 25 13:35:37 1999
From: maxco at telenia.it (maxco)
Date: Thu, 25 Nov 1999 19:35:37 +0100
Subject: Programmatori italiani Python per Mac
Message-ID: <81jvhu$fvk$1@serv1.iunet.it>

Sorry for OT.

Se c'? qualche programmatore di Python su Mac scriva a: tec at daker.com per
una consulenza.







From paul at prescod.net  Sun Nov  7 11:20:36 1999
From: paul at prescod.net (Paul Prescod)
Date: Sun, 07 Nov 1999 10:20:36 -0600
Subject: Python and Smalltalk and Java envy
References: <1270663667-24391282@hypernet.com> <381ED7A6.D7D7FB13@compaq.com> <877lk1wiij.fsf@ev.netlab.co.jp> <19991102154543.A4469@optiplex.palga.uucp> <381F9EB4.63FA033F@python.net> <slrn81ve6m.l1c.neelk@brick.cswv.com> <7vqkqq$oa7$1@nnrp1.deja.com> <382484F1.F8B8D712@home.com>
Message-ID: <3825A6D4.CDFD9E6B@prescod.net>

Wostenberg wrote:
> 
> Intersting observation. Perhaps it comes down to those of us who are
> comfortable with niche languages (Python, Smalltalk, Lisp) and those who
> suffer Java-envy . Do Python people, Ada advocates, or little Lispers wring
> there hands about the fact "we are not the most used language?" Why should
> Smalltalkers?

Programmers of minority languages do often wring their hands for the
simple reason that they are often forced to program in majority
languages.

-- 
 Paul Prescod  - ISOGEN Consulting Engineer speaking for only himself
 http://itrc.uwaterloo.ca/~papresco

"Some of Disney's older properties, such as Peter Pan, Winnie the Pooh,
and Alice in Wonderland, came out of books. But the authors' names are
rarely if ever mentioned, and you can't buy the original books at the
Disney store. If you could, they would all seem old and queer, like very
bad knockoffs of the purer, more authentic Disney versions."
	- http://www.cryptonomicon.com/beginning.html



From stidolph at origin.ea.com  Tue Nov 16 15:01:24 1999
From: stidolph at origin.ea.com (Stidolph, David)
Date: Tue, 16 Nov 1999 14:01:24 -0600
Subject: PythonWin Debugger
Message-ID: <11A17AA2B9EAD111BCEA00A0C9B41793034AAEE8@molach.origin.ea.com>

I have a Win32 app that does its own UI (no MFC available or even standard
windows drawing stuff.  Is there any way to integrate the PythonWin debugger
so it will fire up on an exception or allow me to fire it up?  We are using
the Python DLL.  Advice??

-----Original Message-----
From: Mark Hammond [mailto:mhammond at skippinet.com.au]
Sent: Monday, November 01, 1999 4:58 AM
To: 'Adrian Eyre'; 'Python Mailing List'
Subject: RE: PythonWin Debugger


> > Hmmm.  The debugger should run from all embedded apps,
> except Tkinter
> > ones (and therefore except IDLE).
>
> Does that include straight win32 (non-MFC) apps?

Yes - it should work there.  Eg, if you run
"pythonwin\pywin\debugger\fail.py" as a normal script from Python.exe,
the debugger should fire up correctly.

> > This particular problem would appear to be a bug.  I have another
> > debugger bug I need to track down, and hopefully I can also solve
> > this.
>
> Is it connected with this one, or is it completely different?

Different - occasionally the debugger seems to "hang" (although it
hasnt really - it just doesnt respond - as if a modal dialog is up,
although the modal dialog is hidden).  I can't repro this at the
moment, even tho I have seen it before :-(

> > OTOH, the best stragegy for this is probably remote
> debugging - use a
> > seperate process to perform the debugging.  This would also solve
a
> > few of the hairy "blocking" problems I need to deal with.
However,
> > this doesnt appear trivial either :-(
>
> I should imagine not. I suppose there is some sort of API
> under Windows
> to allow this sort of thing (like the Visual C++ debugger does) but
I
> wouldn't have any idea where to start.

That API is too low-level for Python to use effectively.  We really
dont need the ability to replace an instruction with a break-point.

Mark.


-- 
http://www.python.org/mailman/listinfo/python-list



From fredrik at pythonware.com  Tue Nov 30 08:46:49 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 30 Nov 1999 14:46:49 +0100
Subject: Htmlizing text
References: <943909426snz@vision25.demon.co.uk> <19991129170705.A314@better.net> <19991130141224.A2813@stopcontact.palga.uucp>
Message-ID: <002301bf3b39$5b5bc9e0$f29b12c2@secret.pythonware.com>

Gerrit Holl <gerrit.holl at pobox.com> wrote:
> This is better:
>
> def escapeall(s):
> import htmlentitydefs
> for k, v in htmlentitydefs.items():
> s = string.replace(s, v, '&' + k)
> return s

I assume this is untested code?

> By the way, why doesn't the cgi module does this?

if you use ISO Latin 1 for your pages (which happens to
be default in HTTP/1.1, HTML 3.0, and real life), you
don't really have to quote all possible entities just be-
cause you can...

also see:
http://www.lysator.liu.se/?ttabitars/ (*)

</F>

    "mangling 8-bit data down to 7 bits is broken,
    and has no excuses." -- linus torvalds

*) ie 5.0 users should use this link instead:
http://www.lysator.liu.se/%e5ttabitars/




From mhammond at skippinet.com.au  Tue Nov 30 08:31:51 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Tue, 30 Nov 1999 13:31:51 GMT
Subject: WinNT and python as text filter...
References: <3842f32d.30322451@news.concentric.net>
Message-ID: <blQ04.11116$VA6.52672@news-server.bigpond.net.au>

How are you executing the code?  Are you just typing:

c:\>myscript.py > out

If so, there is a bug in the NT command prompt - you need to specify the
full path to Python.exe on the command line, and it will work...

Mark.

Tim Ottinger wrote in message <3842f32d.30322451 at news.concentric.net>...
>I'm working in WinNT with VIM (good tool) and wanted to write some
>handy filters in python to help me with routine editing chores.
>
>I run my python command at the command line and get the output I want
>printed nicely on my screen. However, if I pipe it to a file the file
>comes out empty, if I pipe it to more I get nothing at all.
>
>Does python know it's in a pipe and act differently? Does 'print' not
>work right in NT? What's up?
>
>My python is:
>
>  PythonWin 1.5.2 (#0, Apr 13 1999, 10:51:12) [MSC 32 bit (Intel)] on
>win32
>  Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>  Portions Copyright 1994-1999 Mark Hammond
>Tim





From python-list at teleo.net  Mon Nov 29 14:03:34 1999
From: python-list at teleo.net (Patrick Phalen)
Date: Mon, 29 Nov 1999 11:03:34 -0800
Subject: Zope 2: External objects?
In-Reply-To: <3842C700.A2685499@gmd.de>
References: <3842C700.A2685499@gmd.de>
Message-ID: <99112911104600.03274@quadra.teleo.net>

[Peter K., on Mon, 29 Nov 1999]
:: Hi, a short question: How do I link external objects,
:: defined in a Python module, into Zope 2?

Once again, this is not the appropriate list for Zope questions.

Go to http://www.zope.org and subscribe to the Zope list and/or
the Zope-Dev list.

However, before posting basic questions like the above to the Zope
lists, *please* read the docs at www.zope.org and at zdp.zope.org,
where you will find the answer you seek.



From bjorn at roguewave.com  Mon Nov  1 14:03:35 1999
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Mon, 1 Nov 1999 12:03:35 -0700 
Subject: Feature request for interactive interpreter
Message-ID: <51C9F3C392D0D111A15600A0C99E64A9AF5606@localhost>

I'm getting tired of typing, so my request for the next version of the
interactive interpreter is a "help" function, probably defined as:

	def help(x): print x.__doc__

(with the appropriate 'no help available' message if the object doesn't have
a doc string...)

pretty-pleas'ly y'rs 
--bjorn



From thomas at bibsyst.no  Wed Nov 24 11:21:29 1999
From: thomas at bibsyst.no (Thomas Weholt)
Date: Wed, 24 Nov 1999 17:21:29 +0100
Subject: CRC-module
References: <38396766.B189D15A@bibsyst.no>
Message-ID: <383C1089.1D0F7C18@bibsyst.no>

Hi,

Ok, so I`ve looked into zlib.crc32 and zlib.adler32. They seem easy
enough to use, but I thought crc-codes had characters and numbers in
them, not just a plain integer like the methods above return. ( As you
can see, I`m a complete ass on this subject, but don`t have time to do
the proper research myself, and was hoping for a "quick fix" ... )

A friend of mine mentioned that I should try SHA-1 instead, for more
accuracy. Can anybody give me an example on how to compute crc-codes,
using zlib or preferrably some more accurate method, for single files ??

If this is all it takes :

crc = module_name.crc_method(file)

and comparison is done like :

if (crc1 == crc2): print "Equal."
else : print "Different."

then all I need is the name of the most effective/accurate module to
use.

If, for some strange reason, I should use one module instead of another,
that info would be interesting too.

I`ll stop bother you now and let you get back to work. ;->

Thomas Weholt



From mwh21 at cam.ac.uk  Tue Nov  9 16:42:59 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 09 Nov 1999 21:42:59 +0000
Subject: what is a function object?
References: <809p0a$8nf$1@nnrp1.deja.com>
Message-ID: <m31z9znxy4.fsf@atrus.jesus.cam.ac.uk>

zakons at my-deja.com writes:

> Hi,
> I use function objects for sorting lists and they are great but I want
> to know more about what a function object is in the python environment.

Well, in one way, a function isn't really much different from an
integer or an instance of a class or any other Python object; they're
builtin objects so you can't derive from them, etc.
 
> * How is a function object different from other python objects? What
> does it include extra? What is missing?

Well, they're of type types.FunctionType. This means that when you try
and call one the interpreter notices that it is a function and does
the right thing.

> * How may function objects be accessed by name in various namespaces?

Exactly as any other data type.

With one caveat: if a function is residing in the __dict__ of a class
object and you access it via . notation, it is automagically wrapped
up into a method object.

> * Other than for sorting, what are typical uses of function objects?

Hmm, passing as arguments to map, callbacks into various libraries,
... oh and calling them.

y'see when the interpreter sees 

     foo(x)

it doesn't know a priori that the name "foo" is bound to a
function. So it looks the name "foo" up, and assuming it finds
something, attempts to call it with argument x. "Attempting to call
it" goes something like:

if type(foo) is types.MethodType:
     apply(foo.im_func,(foo.im_self,x))
elif type(foo) is types.FunctionType:
     apply(foo,(x,))
elif hasattr(foo,'__call__'):
     apply(foo.__call__,(x,))

builtins complicate the above, but I hope you get the idea.

> * What does it mean to say that function objects are first-class
> objects in the python environment?

That you can pass the as arguments, return them from functions and
store them in variables as with any other data type.

> Thanks,
> Stuart Zakon
> Objects by Design
> www.objectsbydesign.com

I hope that helped.
Michael



From 71351.356 at compuserve.com  Wed Nov 17 20:21:18 1999
From: 71351.356 at compuserve.com (T. C. Mits)
Date: Wed, 17 Nov 1999 20:21:18 -0500
Subject: Need help dealing with spaces in NT paths.
References: <80pr76$jq5$1@nnrp1.deja.com> <11f733ec.96958819@usw-ex0101-007.remarq.com> <uogcu93dt.fsf@cs.uu.nl> <80sj4q$l7t$1@nnrp1.deja.com> <80tgi8$hbn$1@ssauraaa-i-1.production.compuserve.com> <uzowdfktm.fsf@cs.uu.nl>
Message-ID: <80vkj6$e20$1@ssauraaa-i-1.production.compuserve.com>

The commands:
os.system(r'dir "Program Files"')
os.system('dir "Program Files"')

work on win98.   Thanks.

However, certain others do not.  Such as:

>>> os.system('"dir" "Program Files"')
Bad command or file name


Though now that I look at it, it seems trivial, some other newbies may also
get tripped up.
Maybe some experts can add the reasons for these kinds of issues in the
Python FAQs (where ever those are).



<piet at cs.uu.nl> wrote in message news:uzowdfktm.fsf at cs.uu.nl...
> >>>>> "T. C. Mits" <71351.356 at compuserve.com> (TCM) writes:
>
> TCM> A simple example of the difficulty is:
> >>>> os.system(r'"dir Program Files"')
> TCM> Bad command or file name
> TCM> 0
> >>>>
>
>
> TCM> This is run at the root of C: for simplicity.
>
>
> TCM> I also get:
>
> TCM> os.popen(r'"dir Program Files"', "r")
> TCM> Bad command or file name
> TCM> <open file '"dir program files"', mode 'r'  at 795930>
>
> TCM> What does that mean?
>
> It means there is no program with the name: dir program files
> You should use:
>
> os.system(r'dir "Program Files"')
> os.popen(r'dir "Program Files"', "r")
> --
> Piet van Oostrum <piet at cs.uu.nl>
> URL: http://www.cs.uu.nl/~piet [PGP]
> Private email: Piet.van.Oostrum at gironet.nl





From skaller at maxtal.com.au  Thu Nov 18 18:32:21 1999
From: skaller at maxtal.com.au (skaller)
Date: Fri, 19 Nov 1999 10:32:21 +1100
Subject: Beginning programming with Python
References: <sng0ychgwt.fsf@motorola.com> <382B3A53.5816B2@callware.com> <382B3D79.74CD3E0F@email.sps.mot.com> <80p9ds$ppo$1@newshost.accu.uu.nl> <asteroidal-80psfo/INN-2.2.1/bat@broadway.news.is-europe.net> <38312D55.A36EB56D@iqsoft.hu> <3831A4EB.76FF32AF@boi.hp.com>
Message-ID: <38348C85.8CC9C61@maxtal.com.au>

Bill Anderson wrote:
> 
> Nemeth Miklos wrote:
> >
> > Janos Blazi wrote:
> >
> > > I have tried to use Perl in my teaching. It has a very powerful but alas!
> > > not very simple syntax! To start with, my pupils used to forget the dollar
> > > signs and Perl does not like that. So this year I shall use Python. We shall
> > > see. Time will tell, if Python is really better. Maybe my pupils will forget
> > > the indenting this time? There is actually nothing else you can forget in
> > > Python.
> >
> > I am quite new in Python programming, and I was a bit :-( by understanding that
> > Python does not have any (even optional) mechanism to enforce the usage of
> > predeclared variables and functions. The complete lack of compile-time
> > type-safety of Pythom may be a problem in large projects, which may be the most
> > serious technical (ie non-marketing) obstacle to Python's becoming a widespread
> > language like Java.
> > Let us see an example:
> > # Module t.py
> > def f():
> >   print "started"
> >   x = 128
> >   x.foo()
> >
> > Let us import it in the Python interpreter:
> > >>> import t
> > >>>
> > You can see thet the interpreter compiled t.py to t.pyc (or t.pyo) but found no
> > error.
> > All typos (typing errors of variable and function names) will be detected ONLY
> > at runtime! Python is perfect language to write programs fast, but with the
> > cost of hordes of testers.
> 
> Disagree. I write many things in python, fast, and without hordes of
> testers.
> 
> > However this may not be a big problem, because any
> > piece of software should be thoroughly tested, to make sure that there are no
> > semantic errors in the application. In the case of Python the applications
> > should also be tested for "syntactic" errors.
> 
> As should _any_ language.
> 
> > In Python (just like in Smalltalk or in Perl, or in Tcl) there are a lot of
> > things a programmer may forget.
[]

-- 
John Skaller, mailto:skaller at maxtal.com.au
10/1 Toxteth Rd Glebe NSW 2037 Australia
homepage: http://www.maxtal.com.au/~skaller
voice: 61-2-9660-0850



From quinn at amatho.ugcs.caltech.edu  Tue Nov 30 20:29:41 1999
From: quinn at amatho.ugcs.caltech.edu (Quinn Dunkan)
Date: 1 Dec 1999 01:29:41 GMT
Subject: What's the canonical vi setting for Python indentation
References: <38445B6A.A1283D2E@cs.mu.oz.au> <Pine.LNX.4.20.9911301540450.30390-100000@seanb.sttln1.wa.home.com> <19991201111751.A1277@Ridcully.home>
Message-ID: <slrn848uhr.o4a.quinn@amatho.ugcs.caltech.edu>

On Wed, 1 Dec 1999 11:17:51 +1100, Malcolm Tredinnick <malcolmt at smart.net.au>
wrote:
>On Tue, Nov 30, 1999 at 11:46:10PM +0000, Sean Blakey wrote:
>> I don't know about canonical, but I have the following in my .vimrc
>> set expandtab	"Turn's tabs into spaces
>> set hardtabs=4	"Make tabs 4 spaces wide
>> set tabstop=4
>> set shiftwidth=4	"For use with << and >>
>
>I tend to avoid 'set expandtab' to save typing. If I want to back out from one
>indentation level to the previous one, it takes a single BackSpace to do so.
>If I am using the 'expandtab' setting, it takes four keystrokes (these things
>matter to me). Of course, I am also using the 'autoindent' setting (but, then
>again, doesn't everybody in vim?).

Note that in vim5, you can set softtabstop=4 to take out expanded tabs with
a single backspace.  Here's my config:

se et ts=8 sw=4 softtabstop=4 smarttab

I like keeping tabs at 8 chars since many other things expect that.
For indentation I use:

se ai
im :<cr>    :<cr><tab>

This mapping also works well for editing natural language, which is a good
example of how intuitive python's indentation rules are :)

>42.7 percent of all statistics are made up on the spot

82.8 percent of those are completely inaccurate.



From reic0024 at ub.d.umn.edu  Tue Nov 16 10:29:03 1999
From: reic0024 at ub.d.umn.edu (Aaron J Reichow)
Date: Tue, 16 Nov 1999 09:29:03 -0600
Subject: Anyway to use file masks with os.listdir()?
Message-ID: <Pine.SO4.4.02.9911160925560.22506-100000@ub.d.umn.edu>

Is there any way to pass a filemask to os.listdir()?  For example, to list
only files with the extension .jpg.  Any help would be appreciated! :)

Aaron




From culliton at clark.net  Wed Nov 24 01:18:39 1999
From: culliton at clark.net (Tom Culliton)
Date: Wed, 24 Nov 1999 06:18:39 GMT
Subject: Returning a variable reference
References: <slrn83jhr5.cl.hutton@el-047.ee.usyd.edu.au> <slrn83mcos.j3.hutton@el-047.ee.usyd.edu.au>
Message-ID: <3rL_3.41286$oa2.253950@iad-read.news.verio.net>

In article <slrn83mcos.j3.hutton at el-047.ee.usyd.edu.au>,
Ross Hutton <hutton at chaffee.usyd.edu.chaffau> wrote:
>Ahh, I think I've just seen the light. Rather than trying to pass an
>lvalue up the chain I should be passing the rrr code down to lower
>access functions (get(rrr) and set(rrr,value)) and have them do the
>dirty work.

Or, since you can pass a tuple anywhere you can pass an index, just
compose and pass (mem, idx) or (reg, idx) around as your destination.
It does take a while to get over the C mentality and realize that
multi-element return values or parameters are both easy and natural in
python.

>Many thanks again to Tom and Ben for their suggestions. You got me talking
>to myself and that's probably not such a bad thing. 
>
>It's quite interesting learning Python which is in many ways a lovely
>language (problem is that I keep thinking in C). I was amazed when I
>did some function pointer work without even looking in a manual. I
>just assigned a function to a variable and called the variable as a
>function. I just wrote code the way I thought it should work (and was
>quite surprised when it did).

More C mindset to unlearn, try to stop thinking of them as pointers.
EVERYTHING in python is passed (and returned) uniformly by reference,
some are immutable like numbers, strings, tuples, functions, others
are mutable like lists, dictionaries, class objects, most extension
types, ...

Maybe you've heard Lisp people talk about functions being "first class
objects"?  This is also true in Python, any "basic" operation you can
do with other types you can do with a function, "assign" it to a
variable, pass it as a parameter, return it, ...

Each type also has their own special operations like indexing sequence
and mapping objects, dumping the keys of a mapping, bitwise operations
on integers, calling functions, ...

It's easy to get fooled by how familiar "assignment" looks, when it's
really a very different mechanism.  Python doesn't have variables as
you know them in C.  A C variable name refers directly to a piece of
storage of a particular size and shape, but in Python it's an indirect
binding to an object of any type.  In C when you say:

x = x + 1

It fetches data from the location refered to by "x" adds 1 to that
value, and stores it back in the same location.  Where in Python, it
looks up "x" in a dictionary, eventually calls the add function with
references to that object and another object containing the integer
value 1, creates a new object containing the result, modifies the
dictionary to point to it, and if the reference count on the object
that x used to refer to has gone to zero garbage collects it.

In C the things you need to internalize to really get it are pointer
math and the central position of expressions.  When you've done this
things like:

char *p, *q; int i;

for (i = 0; i[p] = *(q + i); ++i)
	;

are not suprising, just in bad taste.  To really get Python you need
to internalize the mutable/immutable distinction, and the central
place of dictionaries and references.

Oh well, thats probably way more preaching than you wanted. ;-)

Tom



From tbryan at python.net  Wed Nov 24 06:11:01 1999
From: tbryan at python.net (Thomas A. Bryan)
Date: Wed, 24 Nov 1999 06:11:01 -0500
Subject: Upgrade to version 1.5.2 - won't configure?
References: <Ql2_3.63$Lf2.352159@news.bctel.net> <383990AC.A7D2BD27@python.net> <EEl_3.52$jf7.366030@news.bctel.net> <383A82DE.29ED626F@python.net> <KzH_3.64$Ge5.397561@news.bctel.net>
Message-ID: <383BC7C5.9FCFD75D@python.net>

Don Oliver wrote:

> The Python web site doesn't say that all these extra programs are required,
> nor does the readme in the download.

No.  It assumes that you have a UNIX machine where you can compile C 
programs.  That is, it assumes that you have a functioning C compiler, 
a linker,  appropriate libraries, etc.  Most people developing on a UNIX 
machine don't start using the C compiler by trying to build Python.
If you had decided to build any other program from sources first, then 
you would have encountered this problem then.  What I'm saying is that 
your problem wasn't Python-specific:  it's really more of a Linux/UNIX 
configuration problem.  Perhaps the README should at least mention 
that you must have a functioning C development environment to build 
Python from the sources.

> > Wow!  I never have any idea what's going on with Windows....
>
> I said that I knew what I was doing, I never claimed to know all about the
> DLL's, etc.
> By that, I meant that I had no problem installing and using any programs.
> Big difference!

Just like the rest of us. ;-)  There's nothing wrong with using Python 
under Windows.  I do it often.  Switching to Linux/UNIX, learning Python, 
and learning to compile C programs is a lot to do simultaneously.  

> >Don't we all... seriously, one thing I have learnt over my year or two
> >with Linux is to let rpm do it's stuff and not move stuff around
> >underneath it. It's possible to do it yourself using tarballs and mv
> >and cp, but it becomes harder to keep track of what's where.
> 
> I guess I'll stick with rpms for now.

Good idea. :)  Note that you can still install Python modules such as 
PMW and HTMLgen after you install the basic Python RPMs.  Any Python 
module/package that has no C extension simply requires you to put the 
.py files somewhere where Python can find them.  Many of the most 
commonly used C extensions for Python are available as RPMs, so you 
don't have to build them from the sources either.

but-I-build-all-python-packages-from-sources-on-the-Starship-anyway-ly yours
---Tom



From c.hintze at gmx.net  Sat Nov 27 04:50:55 1999
From: c.hintze at gmx.net (Clemens Hintze)
Date: Sat, 27 Nov 1999 09:50:55 GMT
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <867lj85g4c.fsf@g.local> <al8yabo9fm3.fsf@sirppi.helsinki.fi> <86so1v4jlj.fsf@g.local> <al8hfiaan8p.fsf@sirppi.helsinki.fi> <86ln7m2t31.fsf@g.local> <m3emddrmuh.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <m3emdc1cw0.fsf@qiao.localnet>

You wrote:
    Your> Gareth McCaughan <Gareth.McCaughan at pobox.com> writes:
    Your> > Markus Stenberg wrote:

[...]

    Your> > One quickly gets used to Lisp syntax. Like using
    Your> > indentation to denote block structure. :-) After a while,
    Your> > you just stop noticing all those parentheses.
    Your> 
    Your> This leads to a thought: would it be possible using lisp's
    Your> silly amount of flexibility to get the reader to
    Your> auto-paranthesize (word?  doubt it):
    Your> 
    Your> defun our-member (obj lst)
    Your>   if null lst
    Your>      nil
    Your>      if eql (car lst) obj
    Your> 	  lst
    Your> 	  our-member obj (car lst)

There are several tries to get rid of parentheses. One try is named
LOGO. Your example would be coded in LOGO like that:

to our-member :obj :lst
   ifElse nullp lst
      nil
     ifElse (first lst) == obj
        lst
       our-member obj (first lst)
end 

Okay, I have to admit that indentation is not an issue here ;-)

    Your> indentation-as-structure-*will*-rule-the-world-ly y'rs
    Your> Michael

Dreams-are-allowed-ly y'rs
Clemens



From m.faassen at vet.uu.nl  Thu Nov  4 07:15:13 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 4 Nov 1999 12:15:13 GMT
Subject: Comparison between Python and "Ruby"
References: <1270552684-31022246@hypernet.com> <87n1svugfh.fsf@ev.netlab.co.jp>
Message-ID: <7vrtch$nro$1@newshost.accu.uu.nl>

Yukihiro Matsumoto <matz at netlab.co.jp> wrote:
[snip]
> But TMTOWTDI (in other words: the more, the merrier) is the slogan to
> emphasize maximum variability in ALL AREAS, so that I felt weird to
> see Pythoneers talking TMTOWTDI or even TMTOWTHMTOWTDI in positive
> attitude.  Reducing variablility in ANY AREA contradict against
> TMTOWTDI, I think.

But there are more ways to have more ways to do it, which is I think what
TMTOWTHMTOWTDI stands for. :) Python does it by reducing syntactic variability,
while still having a large semantic variability; and because it's so
easy to use, it arguably has a larger semantic variability than C. Of course
this is nonsense because Python is written in C, but it's just *easier*
to use different programming styles in Python (procedural, object oriented,
functional, etc). More inviting. In this sense Python may even invite
*more* semantic ways to do it than Perl (for instance people sometimes
complain about the OO implementation of Perl).

So there are more ways to have multiple solutions than just Perl's way. As
I'm sure you'll agree with. :)

There-are-more-ways-to-understand-TMTOWTDI-ly yours,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?



From jean-marc.jezequel at irisa.fr  Mon Nov  8 20:43:08 1999
From: jean-marc.jezequel at irisa.fr (Jean-Marc Jezequel)
Date: Mon, 8 Nov 1999 17:43:08 -0800
Subject: Call for Papers - TOOLS Europe 2000
Message-ID: <807u34$dgc$1@news.rain.org>

TOOLS Europe 2000
   "Enterprise Architecture - Patterns - Components"

Mont St Michel & St Malo, France
    June 5-8, 2000

http://www.tools.com/europe


CALL FOR CONTRIBUTIONS (deadline 31st January 2000)

TOOLS is the major international conference series devoted to
applications of object-oriented technology.  TOOLS Europe 2000 will be
held in Le Mont St Michel, at the borderline of Normandy and Brittany
in the north-west of France, and will continue the commitment to
excellence of earlier TOOLS conferences in Europe, Australia, Asia and
the USA.

The proceedings will be published world-wide by the IEEE Computer
Society.


PAPERS

TOOLS Europe 2000 is now soliciting papers on all aspects of
object-oriented technology. All submitted papers will be refereed and
assessed for technical quality and usefulness to practitioners and
applied researchers.

TOOLS Europe particularly welcomes papers that present general
findings based upon industrial experience. Such papers will be judged
by the quality of their contribution to industrial best-practice
(rather than to the body of research knowledge).


TUTORIALS and WORKSHOPS

Tutorials and workshops form an important part of TOOLS
conferences. TOOLS Europe 2000 is keen to receive proposals for
tutorials, workshops and panels on topics.


MORE DETAILS

Please visit the conference website for more details.
http://www.tools.com/europe










From wjk at wjk.mv.com  Sat Nov 27 14:25:04 1999
From: wjk at wjk.mv.com (William J. King)
Date: Sat, 27 Nov 1999 14:25:04 -0500
Subject: yet 'nother newbie question
Message-ID: <3840300F.66DC46DD@wjk.mv.com>

Hi:

When passing an argument to a function etc. -- how does one determine
that it is
an integer that is being passed and not a string for example... Is there
something
available to determine it's type. The only way I can figure is to use
'try' and
'except' for a partial solution:

try:
    x = x + ""    # suceeds then its a string -- if so
   x = string.atoi(x)
except:
    x = x + 0    #This sort of partially ensures that int is passed.....

or if you have the reverse situation?...

Hmmm -- It seems python knows the difference internally, how can I
determine
types and work with variables without having to resort to exception
checking...
Hope this makes sense... Thanks.....








From bciceron at rsn.hp.com  Mon Nov  8 18:13:19 1999
From: bciceron at rsn.hp.com (Benoit Ciceron)
Date: Mon, 8 Nov 1999 17:13:19 -0600 (CST)
Subject: [Python] header (fwd)
Message-ID: <Pine.HPX.4.20.9911081711570.27637-100000@ztt.rsn.hp.com>

can we get an automatic subject header : "[Python]" to ALL messages
submitted to the list , including replies and newsgroup threads ?

thank you,

ben_






From mal at lemburg.com  Fri Nov 12 04:57:33 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Fri, 12 Nov 1999 10:57:33 +0100
Subject: time/strptime/mktime/localtime
References: <Pine.GSO.4.20.9911101632390.1674-100000@sun>
Message-ID: <382BE48D.89CD174B@lemburg.com>

Oleg Broytmann wrote:
> 
> Hello!
> 
>    The following program prints "1999-09-21 21:44:33" on Linux and
> "1999-09-21 22:44:33" on Solaris (1 hour offset).
> 
>    Is it a problem with my config? libs? python? Timezone is Europe/Moscow,
> set (I think) correctly on both systems.
> 
> #! /usr/local/bin/python -O
> 
> import time
> 
> t = time.strptime("1999-09-21 21:44:33", "%Y-%m-%d %T")
> m = time.mktime(t)
> print time.strftime("%Y-%m-%d %T", time.localtime(m))

Could be that one of those sets the DST flag to 0 while
the other sets it to 1... please provide the output of
time.strptime() for both machines.

-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                    49 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/




From tbryan at python.net  Wed Nov 10 21:03:55 1999
From: tbryan at python.net (Thomas A. Bryan)
Date: Wed, 10 Nov 1999 21:03:55 -0500
Subject: ASCII delimited files
References: <942231502.1914033961@news.interbusiness.it>
Message-ID: <382A240B.B247AD6B@python.net>

Roger Irwin wrote:
> 
> Is there any function or module available for parsing ASCII delimited files,
> before I go and re-invent the wheel writing my own.

I'm not sure exactly what you're looking for.  I've appended something 
that I was playing with one day.  It was just a way to create an object 
easily that could parse and validate ascii, delimited files.  
It might be terribly slow: I never timed it.  

Basically, you create a DelimFldParser object with a list of 
DelimParserField subclasses and a delimiter.  Each 
DelimParserField subclass knows how to handle a specific "column" 
of the ASCII file.  The DelimFldParser is then handed a file 
object (anything with a readline() method, really), and it 
returns a list of lists.  The inner list is a list of values 
returned by the DelimFldParser objects for a specific line.
Oh, I also assume that each line of the file has the same number 
of "columns."  

I implemented three sample DelimParserField objects.  One converts 
ascii values to floats.  Another checks that the field value is 
in a specified list of values.  The last is designed to perform 
a verification of field values based on a regular expression.

I wrote this thing to read and verify files before importing them 
into a database.  I never really had much chance to use it, though.
I would love to see someone optimize this thing because it makes the 
task of building a parser for a new format of an ASCII file very 
simple.  It would be great, for example, for dealing with delimited 
data exported from a database or for parsing a delimited file 
for for import into a database.

---Tom


#!/usr/bin/python 

import string
import re

class DelimFldParser:
    def __init__(self, fields, delimiter=None):
        """fields is an ordered list of DelimParserField instances"""
        self.delimiter = delimiter
        self.fields = fields
        self.numCols = len(fields)
        self.cols = []
        for el in fields:
            self.cols.append(el.name)
    def parseLine(self, line):
        list = string.split(line, self.delimiter)
        assert len(list) == self.numCols, \
            "The following line doesn't have enough  fields.\n%s" % line
        for idx in range(self.numCols):
            list[idx] = self.fields[idx].convert(list[idx])
            self.fields[idx].verify(list[idx])
        return list
    def parseFile(self, fileObj):
        data = []
        line = fileObj.readline()
        while line:
            data.append(self.parseLine(line))
            line = fileObj.readline()
        return data
    def __str__(self):
        s = '<DelimFldParser: '
        for el in self.fields:
            s = s + el.name + ', '
        s = s[:-2] + ' >'
        return s

class DelimParserField:
    def __init__(self, name):
        self.name = name
    def convert(self,value):
        return value
    def verify(self,value):
        pass

class EnumField(DelimParserField):
    def __init__(self,name,validValues):
        DelimParserField.__init__(self,name)
        self.validValues = validValues[:]
    def verify(self,value):
        assert value in self.validValues, \
            "%s not in %s on the following line" % (value,self.validValues)

class NumericRngField(DelimParserField):
    def __init__(self,name,start,stop):
        DelimParserField.__init__(self,name)
        self.min = start
        self.max = stop
    def convert(self,value):
        return float(value)
    def verify(self,value):
        assert value >= self.min and value <= self.max, \
          "%s is not between %s an d %s" % (value,self.min,self.max)

class RegexpField(DelimParserField):
    def __init__(self,name,regexp,flags=None):
        DelimParserField.__init__(self,name)
        if flags:
            self.re = re.compile(regexp,flags)
        else:
            self.re = re.compile(regexp)
    def verify(self,value):
        assert self.re.search(value), \
           "%s does not match the pattern '%s'" % (value, self.re.pattern)



if __name__ == '__main__':
    fh = open('delimParser.test','w')
    fh.write("""a 10 9/10/1999
b 3.5 10/11/1974
c 5.7 09/10/1974
""")
    fh.close()
    fh = open('delimParser.test','r')
    myParser = DelimFldParser((EnumField('Enum',('a','b','c')),
                               NumericRngField('Range',0,10),
                               RegexpField('RegExp','\d{1,2}/\d{2}/\d{4}')))
    print myParser
    output = myParser.parseFile(fh)
    fh.close()
    print output



From olipt at mayo.edu  Tue Nov  9 17:14:58 1999
From: olipt at mayo.edu (Travis Oliphant)
Date: Tue, 9 Nov 1999 16:14:58 -0600
Subject: PyArg_ParseTupleAndKeywords problem
In-Reply-To: <38287096.689640FD@ioc.ee>
References: <38287096.689640FD@ioc.ee>
Message-ID: <Pine.LNX.4.10.9911091613220.12332-100000@us2.mayo.edu>

> 
> So, I write the following C code:
> <snip>
> static PyObject *test_foo(PyObject *capi_self, PyObject *capi_args,
> PyObject *capi_keywds) {
>     PyObject *a_capi = NULL;
>     PyObject *b_capi = NULL;
>     static char *capi_kwlist[] = {"b",NULL};

When using PyArg_ParseTupleAndKeywords I've always had to specify a
keyword string for ALL of the arguments:

     static char *capi_kwlist[] = {"a","b"};

> So, clearly PyArg_ParseTupleAndKeywords command emulates only a subset
> of Python possibilities.


Travis





From claudius at catlover.com  Wed Nov 24 16:16:54 1999
From: claudius at catlover.com (claudius at catlover.com)
Date: Wed, 24 Nov 1999 21:16:54 GMT
Subject: preprocessor for Python?
Message-ID: <aBY_3.23149$V4.277574@news1.frmt1.sfba.home.com>

All, I'm interested in a CPP-style macro preprocessor for Python development...
Are there any such beasts?

Primarily, I'm looking for some macro facilities so I can include debug
statements that are removed when I need to test performance.  (And yes I know
of the assert statement...)



From pnagel at tbs.co.za  Thu Nov 18 05:24:42 1999
From: pnagel at tbs.co.za (Pieter Nagel)
Date: Thu, 18 Nov 1999 12:24:42 +0200
Subject: Beginning programming with Python
In-Reply-To: <assonant-80s673/INN-2.2.1/basilar@broadway.news.is-europe.net>
References: <sng0ychgwt.fsf@motorola.com> <382B3A53.5816B2@callware.com> <382B3D79.74CD3E0F@email.sps.mot.com> <80p9ds$ppo$1@newshost.accu.uu.nl> <asteroidal-80psfo/INN-2.2.1/bat@broadway.news.is-europe.net> <38312D55.A36EB56D@iqsoft.hu> <antigorite-80rol2/INN-2.2.1/attentive@broadway.news.is-europe.net> <19991116173142.A31379@optiplex.palga.uucp> <assonant-80s673/INN-2.2.1/basilar@broadway.news.is-europe.net>
Message-ID: <Pine.LNX.4.10.9911181223450.1914-100000@iguana.home.tbs.co.za>

On Tue, 16 Nov 1999, Janos Blazi wrote:

> From: Janos Blazi <jblazi at netsurf.de>

> > the less time it costs to program it, the more time it costs to run it,
> > the more time it costs to program it, the less time it costs to run it.
> 
> Is this some law of nature I missed at university?
> But obviously, as far as raw speed is concerned Python cannot compete with
> C. Nobody can attempt to deny that. So if speed is an issue for you, use C.

But then of course 90% of the time is spent in only 10% of the code.
So I'd write 90% of the code in Python and just the speed-critical
bit in C.

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

    Tangible Business Software
    http://www.tbs.co.za
    +27 12 46 8563	




From tbryan at python.net  Mon Nov 15 07:08:42 1999
From: tbryan at python.net (Thomas A. Bryan)
Date: Mon, 15 Nov 1999 07:08:42 -0500
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <JlFX3.3659$s87.67930@dfiatx1-snr1.gtei.net>
Message-ID: <382FF7CA.319CCC1@python.net>

Jules Allen wrote:

> The best part of your story for me is confirmation that open source
> projects don't do well without strong leadership. No ESR or an
> ESR-like figure to crack the whip, no Trove.

In my experience, strong leadership is only necessary to maintain 
direction in an Open Source project.  To get work done, you don't 
need a leader cracking the whip; you need one or more programmers 
who really want to solve some problem.  As long as they are 
motivated, the project will progress.  If they are solving a 
problem that plagues or interests other programmers, then others 
will volunteer to contribute.  (This is where leadership comes in...
keeping all interested contributors moving in the same direction.)  
It progresses this way until all of the current maintainers get 
bored and move on...but since the source is open, others can pick 
it up and use it to start new development.

if-guido-got-hit-by-a-bus,-python-wouldn't-die-ly yours

---Tom



From mlh at vier.idi.ntnu.no  Tue Nov 16 04:20:15 1999
From: mlh at vier.idi.ntnu.no (Magnus L. Hetland)
Date: 16 Nov 1999 10:20:15 +0100
Subject: list.without()?
References: <NDBBIAPHEKGOBIHCBLIICEOMCAAA.mcfletch@vrtelecom.com>
Message-ID: <y0jiu32iyi8.fsf@vier.idi.ntnu.no>

"Mike Fletcher" <mcfletch at vrtelecom.com> writes:

> Well...
> 
> def without( source, element):
> 	temp = source[:]
> 	while temp:
> 		try:
> 			temp.remove( element )
> 		except:
> 			break
> 	return temp

Whoa!

Why on *earth* would you use a while/try/break-combination? IMO a much
more natural solution would be:

def without(source, element):
    result = source[:]
    if element in result:
        result.remove(element)
    return result

My point was that I wanted a method in the standard libraries, or,
preferrably as a list method. I guess the closest I got wat the filter
suggestion.

--

  Magnus          Echelon jamming noise:
  Lie             FBI CIA NSA Handgun Assault Bomb Drug Terrorism
  Hetland         Special Forces Delta Force AK47 Hillary Clinton 



From aa8vb at yahoo.com  Wed Nov 17 09:02:14 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Wed, 17 Nov 1999 09:02:14 -0500
Subject: wxPython and DISLIN work nicely together
In-Reply-To: <80s0b4$g5n@mail.psy.uva.nl>
References: <80s0b4$g5n@mail.psy.uva.nl>
Message-ID: <19991117090214.B807152@vislab.epa.gov>

Ionel Simionescu:
 |DISLIN (http://www.linmpi.mpg.de/dislin/) is a powerful visualization
 |package available on quite a lot of platforms, aimed at scientists.
 |DISLIN is "ported" to Python as well.
 |
 |I want to report that wxPython and DISLIN work very well together.

If only wxPython would build on IRIX...

-- 
Randall Hopper
aa8vb at yahoo.com



From m.faassen at vet.uu.nl  Thu Nov 25 04:00:31 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 25 Nov 1999 09:00:31 GMT
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <81g8po$reo$1@mach.vub.ac.be> <81h2hh$85c$1@nntp9.atl.mindspring.net>
Message-ID: <81itrf$eqc$1@newshost.accu.uu.nl>

Aahz Maruch <aahz at netcom.com> wrote:
> In article <81g8po$reo$1 at mach.vub.ac.be>,
> Thomas Hamelryck <thamelry at vub.ac.be> wrote:
>>
>>Block delimitation by indenting. This makes python look like a porsche
>>with a caravan attached :-). As far as I know, most python users even
>>seem to like this for some reason.

> I agree with you and I'm also one of the people who likes it.  Here's
> why: it makes Python code *readable*.  In fact, I wish that Guido would
> disallow

>   if foo : bar()

> and force it to always be

>   if foo :
>     bar()

Agreed. The Zope sources are full of the former and that idiom doesn't help
me read them particularly.

Perhaps 1.6 could introduce this requirement, and some scripts to convert
existing code. The only use I can see for the former is to write nice
obfuscated one liners. :)

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?



From pinard at iro.umontreal.ca  Mon Nov  1 20:01:01 1999
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 01 Nov 1999 20:01:01 -0500
Subject: Default indentation of 4 columns
In-Reply-To: Michael Hudson's message of "24 Oct 1999 18:42:36 +0100"
References: <oq904sc08y.fsf@titan.progiciels-bpi.ca> <m3904s1wqb.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <oq7lk1hfle.fsf@titan.progiciels-bpi.ca>

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

> It hadn't occurred to me before, but one `solution' would be to chuck
> another space between `if' and `(', like so:

> def r():
>     if  (long_condition
>          and other_condition):
>         return ineteresting_stuff

> Not especially natural, but it might just work...

Yes.  Not bad! :-)  Thanks for the suggestion.

P.S. - Thanks to Emile van Sebille <emile at fenx.com> for sharing his idea:

> How about :

>     for file_name, line_number, function_name, python_text in frames:
>         if (function_name[0] not in ('_', '<')
>             and function_name not in transparent_functions
>             ):
>             name = os.path.basename(file_name)
>             if name not in transparent_files:
>                 return name, line_number

> At least here, the '):' is clear.

It was the next best suggestion (not raising fights with Python Emacs mode).
I did try it a few times, and I'm not sure I'll easily tame myself to it.

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





From aahz at netcom.com  Fri Nov  5 18:45:07 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 5 Nov 1999 23:45:07 GMT
Subject: What are Aspects? (Re: How can I make my assertions smarter?)
References: <Pine.WNT.4.05.9911032042440.109-100000@david.ski.org> <slrn824lqc.ngk.wtanksle@hawking.armored.net>
Message-ID: <7vvq63$ehm$1@nntp4.atl.mindspring.net>

In article <slrn824lqc.ngk.wtanksle at hawking.armored.net>,
William Tanksley <wtanksle at hawking.armored.net> wrote:
>
>In short, aspects are like objects, only where an object implements a
>self-contained concept, an aspect implements a widespread concept.  For
>example, reference counting could be written as an aspect, but never as an
>object.

Sounds like templates to me, or Python's "mix-in" classes.  Am I missing
something?
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From hildeb at www.stahl.bau.tu-bs.de  Fri Nov 26 11:22:41 1999
From: hildeb at www.stahl.bau.tu-bs.de (Ralf Hildebrandt)
Date: 26 Nov 1999 16:22:41 GMT
Subject: fcntl/ioctl and O_NONBLOCK
References: <slrn83t5mv.bhk.hildeb@stahlw06.stahl.bau.tu-bs.de>
Message-ID: <slrn83tcuh.doh.hildeb@stahlw06.stahl.bau.tu-bs.de>

On 26 Nov 1999 14:19:11 GMT, Ralf Hildebrandt <hildeb at www.stahl.bau.tu-bs.de> wrote:

>What I want to do: I want to use select() to see if there's data to be
>read from stdin or stdout and read it ALL (thus I need stdin/stdout to
>be non blocking!)

Hmm, is this here correct ?

fcntl.fcntl(stdout.fileno(), FCNTL.F_SETFD, FCNTL.O_NDELAY)



From robin.boerdijk at nl.origin-it.com  Wed Nov 10 11:14:09 1999
From: robin.boerdijk at nl.origin-it.com (Robin Boerdijk)
Date: Wed, 10 Nov 1999 17:14:09 +0100
Subject: .dll and .exe file version informaton
References: <000101bf2b3c$a7d43e40$d32d153f@tim>
Message-ID: <0E16861EE7BCD111BE9400805FE6841F0C7A489C@c1s5x001.cor.srvfarm.origin-it.com>

Tim Peters <tim_one at email.msn.com> wrote in message
news:000101bf2b3c$a7d43e40$d32d153f at tim...
> [emuller at painewebber.com]
> > Can anyone give me an idea how to quikly get the version info on .dll
> > and .exe files under windows?
>
> In Explorer, right-click on the file, select Properties from the context
> menu, and click the Version tab.

Strange... My Explorer only gives me a General tab. Quickview in the context
menu, however, shows a lot of technical info about the selected .dll or
.exe.

Robin






From nick at src.uchicago.edu  Wed Nov  3 15:38:42 1999
From: nick at src.uchicago.edu (Nick Collier)
Date: Wed, 03 Nov 1999 14:38:42 -0600
Subject: Can't start IDLE
References: <381F4EB1.AFD9BA32@src.uchicago.edu> <5lg0yotka6.fsf@eric.cnri.reston.va.us>
Message-ID: <38209D52.C4370040@src.uchicago.edu>

It looks like this is the problem. Unfortunately, some other software
I've got installed relies on other versions of tcl, so I'll have to run
emacs as my python editor for now. Saw IDLE at the talk you gave at the
Univ. of Chicago last week and wanted to give it a try. I enjoyed the
talk and thanks for coming.

Nick

Guido van Rossum wrote:
> 
> Nick Collier <nick at src.uchicago.edu> writes:
> 
> > On winNT sp3 I get the following message when I do a "python idle.py"
> > from the command line:
> >
> > C:\Program Files\Python\Tools\idle>python idle.py
> > Traceback (innermost last):
> >   File "idle.py", line 3, in ?
> >     PyShell.main()
> >   File "PyShell.py", line 612, in main
> >     fixwordbreaks(root)
> >   File "EditorWindow.py", line 606, in fixwordbreaks
> >     tk.call('tcl_wordBreakAfter', 'a b', 0) # make sure word.tcl is
> > loaded
> > TclError: invalid command name "tcl_wordBreakAfter"
> >
> > I looked around deja.com and found one other person with the same
> > problem, but no response so I'm asking here.
> 
> Could it be that you have an older (or newer) Tcl version installed?
> Python (for now, until I upgrade the _tkinter module) requires Tcl/Tk
> 8.0.5 -- which is the version that comes with the Python 1.5.2 for
> Windows installer.
> 
> Try to erase all traces of other Tcl installs and reinstall Python and
> see if that works.
> 
> --Guido van Rossum (home page: http://www.python.org/~guido/)

-- 

Nick Collier
nick at src.uchicago.edu
Social Science Research Computing
University of Chicago
Chicago, IL 60637



From duncan at rcp.co.uk  Wed Nov 10 07:01:17 1999
From: duncan at rcp.co.uk (Duncan Booth)
Date: 10 Nov 1999 12:01:17 GMT
Subject: Newbee WinNT ODBC Question
References: <809ckk$vbf$1@nnrp1.deja.com>
Message-ID: <8E7A7ABCDduncanrcpcouk@news.rmplc.co.uk>

wayne_strauss at mail.medscape.com wrote in <809ckk$vbf$1 at nnrp1.deja.com>:

>I've installed Python and Win32all.exe. When I run the following (only
>the first three lines shown):
>
>import dbi, odbc     # ODBC modules
>import time          # standard time module
>
>dbc = odbc.odbc('db/id/pw')
>
>I get:
>
>>>> Traceback (innermost last):
>  File "d:\Program
>Files\Python\Pythonwin\pywin\framework\scriptutils.py", line 237, in
>RunScript
>    exec codeObject in __main__.__dict__
>  File "D:\dev\Python\ODBC\odbc.py", line 1, in ?
>    import dbi, odbc     # ODBC modules
>  File "odbc.py", line 4, in ?
>    dbc = odbc.odbc('db/id/pw')
>TypeError: call of non-function (type module)
>
>If the import worked then why does the actual function call fail?

It looks rather as though you called your file odbc.py so when you do 
import odbc it tries to reimport your script instead of importing the odbc 
module. Rename your source file and it should work a bit better.





From philh at welshnet.co.uk  Mon Nov 29 17:53:07 1999
From: philh at welshnet.co.uk (Phil Harris)
Date: Mon, 29 Nov 1999 22:53:07 -0000
Subject: Exposing COM via XML-RPC or Something Else
References: <613145F79272D211914B0020AFF6401914DD7C@gandalf.digicool.com> <00f001bf3b6c$5debe270$4500a8c0@thomasnotebook>
Message-ID: <944002498.101563@zx81.mersinet.co.uk>

Thomas Heller <thomas.heller at ion-tof.com> wrote in message
news:00f001bf3b6c$5debe270$4500a8c0 at thomasnotebook...
> > You might want to take a look at MS' SOAP (Simple Object Access
> > Protocol) spec - it's not fully baked yet, but it basically takes
> > xml-rpc several steps further, allowing you to communicate with
> > SOAP-enabled objects (COM or otherwise) using XML over HTTP. We in
> > Zope-land are keeping a close eye on this, as publishing objects
> > via HTTP is the core idea of Zope already.
>
> Zope and SOAP sounds very similar.
> Is this by accident?

yes

>
> Thomas Heller
>
>
>





From janne at nnets.fi  Thu Nov 11 07:04:01 1999
From: janne at nnets.fi (Janne Sinkkonen)
Date: 11 Nov 1999 14:04:01 +0200
Subject: How do I compile NumPy on Linux?
References: <382A8094.3354DE70@mdmlab.mi.infm.it> <19991111064736.A528@better.net>
Message-ID: <m3n1sl1bgu.fsf@kinos.nnets.fi>

William Park <parkw at better.net> writes:

> On Thu, Nov 11, 1999 at 09:38:45AM +0100, Sandro Ferrari wrote:
> > I downloaded the LLNLDistribution.tgz file and followed the installation
> > instruction:
> > cd Numerical
> > ./makethis.py
> > but I get the message:
> > make: Makefile.pre.in: No such file or directory
> > make: *** No rule to make target `Makefile.pre.in'.  Stop.
> > make: *** No targets.  Stop.
> > In the documentation isn't mentioned anything about it.
> > Any help?
> > thanks
> > Sandro
> 
> This was long time ago.  But, from my memory, I did
>     python makethis.py
>     python installthis.py
> So, check the *.py files whether this should be.

It's identical to what he tried. I guess that he either has an
incomplete Numeric package (unlikely) or that
/usr/lib/python1.5/config/Makefile.pre.in or an equivalent is
missing. I don't know whether the config directory is provided by the
python RPM or by the python-devel RPM, but the config directory _was_
missing from a Python 1.5.2. RPM which I installed some months ago
(all this assumes he uses RedHat). A suitable combination of RPMs or
installing from sources fixes this problem.

-- 
Janne



From pmoran at nas.nasa.gov  Tue Nov  2 17:35:34 1999
From: pmoran at nas.nasa.gov (Patrick J. Moran)
Date: 2 Nov 1999 22:35:34 GMT
Subject: NT DLL modules and C++
Message-ID: <7vnovm$6do$1@sun500.nas.nasa.gov>

Hello,

I am using Python as a wrapper around several C++ libraries that
our group has developed.  My approach is to make dynamically linked
modules and then import them.  The libraries contain classes with
static data members that need to be initialized when the module is
loaded.  On SGI's, if I build Python with "LDSHARED=CC -n32 -shared -all",
then Python knows to call the C++ initialization routine when my module
is loaded.  How do I get Python to call the initialization under NT?

Thanks.

Pat

# --------------------------------------------------------------------
# Patrick J. Moran                     http://www.nas.nasa.gov/~pmoran
# Veridian/MRJ Technology Solutions          NASA Ames Research Center



From mstenber at cc.Helsinki.FI  Fri Nov 26 10:46:24 1999
From: mstenber at cc.Helsinki.FI (Markus Stenberg)
Date: 26 Nov 1999 17:46:24 +0200
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <867lj85g4c.fsf@g.local> <383CC835.B585EBDD@Lugoj.Com> <m3ogcjrn41.fsf@atrus.jesus.cam.ac.uk> <383D9C63.3A7E5806@Lugoj.Com> <86ogci2tfm.fsf@g.local>
Message-ID: <al8g0xt5k8f.fsf@sirppi.helsinki.fi>

Gareth McCaughan <Gareth.McCaughan at pobox.com> writes:
> James Logajan wrote:
> > Oh yeah? Well yer mudder wears army boots! Seriously though, the price I am
> > talking about has absolutely nothing to do with simplicity. It has to do
> > with the fact that languages such as Java that have "true" GC are
> > non-deterministic in their execution times because GC can cause execution
> > pauses in the VM at unpredictable times. This makes Java unacceptable for a
> > large number of real-world applications that I have been involved in lately.
> > ("I'm sorry, but your 911 call can not go through for another 20 seconds
> > while our Java VM garbage collects 1GB of memory on the switch.") I have
> > recommended Python over Java in at least one environment because of Python's
> > deterministic approach to memory management. Had Python not had that, I
> > suppose we would have gone with some other scripting language or even
> > designed our own (gak!).
> Garbage collectors have got a *lot* better since the days when
> GC meant half-second pauses waiting for the machine to sweep
> through all its memory. Let alone 20 seconds.

Where can I admire such a beast? At least none of the Java
implementations(*) I have tried(**) have one. say, for 128MB of trash, it's
consistent N second sweep (0.5-10, depending on hardware).

> -- 
> Gareth McCaughan  Gareth.McCaughan at pobox.com
> sig under construction

(*)
	HPUX JVM, Linux Sun/IBM JVM, Windows Sun JVM

(**)
	I could have tried harder. Yes. But I'm hoping for clues here <0.5wink>
-- 
The IBM Principle:		  
	Machines should work.  People should think.
The Truth About the IBM Principle:
	Machines don't often work, people don't often think.



From krutz at cg.tuwien.ac.at  Wed Nov 24 13:57:02 1999
From: krutz at cg.tuwien.ac.at (Markus Krutz)
Date: Wed, 24 Nov 1999 18:57:02 GMT
Subject: newbie question
Message-ID: <943469319.838266356@news.teleweb.at>

How the f*** do I create a string containing a *single* "\" ?

regards
Markus




From mcfletch at vrtelecom.com  Mon Nov 29 08:35:30 1999
From: mcfletch at vrtelecom.com (Mike Fletcher)
Date: Mon, 29 Nov 1999 08:35:30 -0500
Subject: [wxPython] Mandelbrot - further improvements
In-Reply-To: <NDBBIAPHEKGOBIHCBLIICEAOCBAA.mcfletch@vrtelecom.com>
Message-ID: <NDBBIAPHEKGOBIHCBLIIMEAOCBAA.mcfletch@vrtelecom.com>

Urgh,  forgot the attachment... here it is...

-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: mandelbrot.py
URL: <http://mail.python.org/pipermail/python-list/attachments/19991129/599be2b0/attachment.ksh>

From Josef.Betancourt at GTECH.COM  Wed Nov  3 11:23:24 1999
From: Josef.Betancourt at GTECH.COM (Betancourt, Josef)
Date: Wed, 3 Nov 1999 11:23:24 -0500 
Subject: capturing stderr && stdout
Message-ID: <7780536C9FD5D211801B0008C75D433933D00D@rimail03>

Agent:
I'm using:
    if on98 :
        program = os.popen(cmdl, 'r')
    elif onNT :
        program = os.popen('( ' + cmdl + ' ) 2>&1', 'r')
    else:   # this assumes Solaris...
        program = os.popen('{ ' + cmdl + '; } 2>&1', 'r')
    
It works,but stdout and stderr are both captured.

-- Josef

> -----Original Message-----
> From: Agent Drek [mailto:drek at MonsterByMistake.Com]
> Sent: Tuesday, November 02, 1999 2:19 PM
> To: python-list at python.org
> Subject: capturing stderr && stdout
> 
> 
> Hullo,
> 
> I have searched the web and all the python documents/examples 
> I could find
> before posting here. My apologies if I have missed the 
> obvious :) (I bought
> two books as well that I am going through)
> 
> Alot of what I need to do with python depends on launching 
> and controlling
> a subprogram called hscript. It's a 'closed' application that 
> is not a full
> bodied language like python. I am hoping to launch one 
> instance of hscript
> write to it and keep its stdout && stderr.
> 
> Most of our work gets done with csh scripts like the following:
> <csh_script>
> #!/bin/csh -f
> 
> #csh stuff goes here
> 
> hscript -q $HIPDIR/$HIPFILE << ENDCAT >>&! $LOG
> 	# now I write hscript commands
> 	opcf /obj
> 	#etc
> ENDCAT
> exit
> </csh_script>
> 
> here's a stripped down version of what I have thus far.
> 
> <python_script>
> #!/usr/local/bin/python
> 
> import os, sys
> HSCRIPT = "hscript"
> 
> p = os.popen("%s" % HSCRIPT, "w")
> p.write("opcd /obj\n")
> p.write("opadd geo\n")
> p.write("opls\n")
> p.write("quit\n")
> </python_script>
> 
> This is great so far but I just need stderr from hscript 
> (just in case it
> crashes:) and stdout (to keep a neat log file)
> 
> This will need to work on *nix.
> 
> I'm trying to move to python as our main scripting language 
> so that we can
> reuse code and integrate a gui (tkinter) easily into our apps.
> 
> If you're interested in what hscript is you can visit 
http://www.sidefx.com/
and check out houdini.

thanks for any tips,

	=derek

Monster By Mistake Inc > 'digital plumber'
mailto:drek at monsterbymistake.com
http://www.interlog.com/~drek
"This mail was powered by Linux & Vim"


-- 
http://www.python.org/mailman/listinfo/python-list




From darcy at vex.net  Thu Nov 11 09:40:43 1999
From: darcy at vex.net (D'Arcy J.M. Cain)
Date: 11 Nov 1999 14:40:43 GMT
Subject: ASCII delimited files
References: <942231502.1914033961@news.interbusiness.it>
Message-ID: <80ekhb$1qrh$1@hub.org>

Roger Irwin <irwin at mail.com> wrote:
> Is there any function or module available for parsing ASCII delimited files,
> before I go and re-invent the wheel writing my own.

> BTW, I know it is simple to do, but I was thinking along the lines that a
> 'standard' could easily be substituted for a C version for large files.

I use delimited.py which works pretty good.  Here's an example usage.

import delimited
for l in delimited.split_delimited(open('datafile.csv').read()):
    print "Field 1: %s\nField 2: %s\n\n" % (l[0], l[1])

I couldn't find this on the Python or Starship sites but you can get it
from http://www.deja.com/getdoc.xp?AN=448769706.  One small problem with
it is that it doesn't handle commas within quoted strings properly.  I
am going to look at this today for another project and I will post it
here when I have something.

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



From tlane at parlant.com  Mon Nov  8 12:17:10 1999
From: tlane at parlant.com (Thomas Lane)
Date: Mon, 08 Nov 1999 10:17:10 -0700
Subject: Image buttons with Tkinter
Message-ID: <38270596.A391995C@parlant.com>

I'm trying to create a button that displays an image instead of text,
but I don't seem to be having any luck. The button just comes up gray.
Here is a code snippet of what I'm trying to do:

		# Put remove.gif on a button (actionbar is a frame for various
controls)
		deletebutton = Button(actionbar,image=PhotoImage(file="remove.gif"))
		deletebutton.place(x=0)

What am I doing wrong here?

Also, is there a place on the net where I can find lots of sample
Python/Tkinter code? I've read Fredrik Lundh's intro, but I haven't been
able to find any other code or reference material for Tkinter.

-- 
Thomas Lane
Provo, UT
tlane at parlant.com



From mhammond at skippinet.com.au  Thu Nov 11 00:39:16 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 11 Nov 1999 05:39:16 GMT
Subject: ASCII delimited files
References: <942231502.1914033961@news.interbusiness.it> <382A240B.B247AD6B@python.net> <382A2801.FDE3CA09@python.net> <382A4330.833C4EBD@easystreet.com>
Message-ID: <8EsW3.962$Xo.4938@news-server.bigpond.net.au>

Al Christians wrote in message <382A4330.833C4EBD at easystreet.com>...
>There's a trick with files of delimited fields.  If a string field
>contains a delimiter (e.g. a comma in a csv file), then the field
>gets enclosed in quotes.  The handling of quotes is probably trickier,
>IDK what the rules are if a field starts with or contains a single or
>double quote mark.  I think I remember that some spreadsheets used tp
>routinely enclose all strings in quotes in CSV files, but I'm not sure.


Yes - a good CSV parser would make a very useful addition to the library!

Mark.





From Jorrit.Tyberghein at uz.kuleuven.ac.be  Wed Nov 24 04:57:08 1999
From: Jorrit.Tyberghein at uz.kuleuven.ac.be (Jorrit Tyberghein)
Date: Wed, 24 Nov 1999 10:57:08 +0100
Subject: Problem installing Python 1.5.2 on Solaris
Message-ID: <943437439.63660@marvin>

Hi,

I'm trying to compile Python 1.5.2 on Solaris.
uname -a gives the following info:
    SunOS poirot 5.6 Generic_105181-15 sun4u sparc SUNW,Ultra-2

I do a configure which seems to work ok and then I type 'make'. This works
for a while but then I get the following errors:

gmake[1]: Entering directory `/users/poirot/c3t2d0s0/jorrit/varia/engine/Python-1.5.2/Modules'
gcc  -g -O2 -I./../Include -I.. -DHAVE_CONFIG_H -c ./regexmodule.c
gcc  -g -O2 -I./../Include -I.. -DHAVE_CONFIG_H -c ./regexpr.c
gcc  -g -O2 -I./../Include -I.. -DHAVE_CONFIG_H -c ./pcremodule.c
gcc  -g -O2 -I./../Include -I.. -DHAVE_CONFIG_H -c ./pypcre.c
gcc  -g -O2 -I./../Include -I.. -DHAVE_CONFIG_H -c ./posixmodule.c
In file included from ./posixmodule.c:3031:
/usr/include/sys/statvfs.h:38: parse error before `fsblkcnt_t'
/usr/include/sys/statvfs.h:38: warning: no semicolon at end of struct or union
/usr/include/sys/statvfs.h:39: warning: data definition has no type or storage class
/usr/include/sys/statvfs.h:40: parse error before `f_bavail'
/usr/include/sys/statvfs.h:40: warning: data definition has no type or storage class
/usr/include/sys/statvfs.h:41: parse error before `f_files'
/usr/include/sys/statvfs.h:41: warning: data definition has no type or storage class
/usr/include/sys/statvfs.h:42: parse error before `f_ffree'
/usr/include/sys/statvfs.h:42: warning: data definition has no type or storage class
/usr/include/sys/statvfs.h:43: parse error before `f_favail'
/usr/include/sys/statvfs.h:43: warning: data definition has no type or storage class
/usr/include/sys/statvfs.h:51: parse error before `}'
/usr/include/sys/statvfs.h:51: warning: data definition has no type or storage class
./posixmodule.c: In function `posix_fstatvfs':
./posixmodule.c:3044: storage size of `st' isn't known
./posixmodule.c: In function `posix_statvfs':
./posixmodule.c:3096: storage size of `st' isn't known
gmake[1]: *** [posixmodule.o] Error 1

Hope you can help me.

Greetings,


--
---------------------------------------------------------------------------
Jorrit Tyberghein (Jorrit.Tyberghein at uz.kuleuven.ac.be)
Project Manager of Crystal Space, a free and portable 3D 6DOF portal engine.
More info at: http://crystal.linuxgames.com
---------------------------------------------------------------------------






From fredrik at pythonware.com  Thu Nov 25 03:58:13 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 25 Nov 1999 09:58:13 +0100
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <867lj85g4c.fsf@g.local> <383CC835.B585EBDD@Lugoj.Com>
Message-ID: <00bd01bf3723$36d699c0$f29b12c2@secret.pythonware.com>

James Logajan <JamesL at Lugoj.Com> wrote:
> Sorry, but I think the price paid for "true" GC is too high for the small
> number of programs that benefit from this. Perhaps one should learn to clean
> up after oneself when cycles are expected.

Hmm.  I've recently dissected two large (and very complicated)
Python projects, and one of the major cost contributors when
going from "working code" to "production quality code" was the
time needed to hunt down a number of totally unexpected re-
source leaks.  GC would have helped in most of those cases.

(and in contrast what people claim on this list, more
syntactic sugar or static typing would have saved al-
most nothing...)

I plan to write a longer piece on this at a later time.

</F>




From hamish_lawson at yahoo.co.uk  Tue Nov 16 05:44:51 1999
From: hamish_lawson at yahoo.co.uk (hamish_lawson at yahoo.co.uk)
Date: Tue, 16 Nov 1999 10:44:51 GMT
Subject: Python module to format text into columns?
References: <80oq13$ql3$1@nnrp1.deja.com> <y0jogcvic4t.fsf@vier.idi.ntnu.no>
Message-ID: <80rcj3$n23$1@nnrp1.deja.com>

mlh at vier.idi.ntnu.no (Magnus L. Hetland) wrote:

> Without really knowing whether there is such a module, I offer a
> simple implementation of some of the functionality, which can easily
> be extended:

Holding to the principle of not reinventing the wheel, I decided to
search the Python archives and (if necessary) ask in this newsgroup
before attempting to write my own module. Now that it appears that no
such module already exists, your contribution, Magnus, has given me a
start in writing my own. Thank you.

The module could provide a format picture mechanism for specifying
column widths and alignments:

   formatter.setpicture("    <<<<<<<<<<<<<<<<  <<<<  >>>>>>")

but I'm inclined to provide (at least initially) just the direct
methods:

   formatter.setcolumn(column=1, margin=4, width=16, align=left)
   formatter.setcolumn(column=2, margin=2, width=4, align=left)
   formatter.setcolumn(column=3, margin=2, width=6, align=right)

Does anyone have any thoughts on this?

Hamish Lawson


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



From phd at phd.russ.ru  Wed Nov 24 10:47:03 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Wed, 24 Nov 1999 15:47:03 +0000 (GMT)
Subject: CRC-module
In-Reply-To: <383C1089.1D0F7C18@bibsyst.no>
Message-ID: <Pine.LNX.4.21.9911241545510.4650-200000@fep132.fep.ru>

On Wed, 24 Nov 1999, Thomas Weholt wrote:
> crc = module_name.crc_method(file)
> 
> and comparison is done like :
> 
> if (crc1 == crc2): print "Equal."
> else : print "Different."
> 
> then all I need is the name of the most effective/accurate module to
> use.

   Use md5. It is one of the best in this area. Attached module has a
method that you can use to md5 a file.

Oleg.
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.
-------------- next part --------------
#! /usr/local/bin/python
"""
   User wrapper for md5 builtin object

   Written by BroytMann, Jun 1997. Copyright (C) 1997 PhiloSoft Design
"""

from md5 import md5

class md5wrapper:
   def __init__(self, init=None):
      if init:
         self._md5 = md5(init)
      else:
         self._md5 = md5()

   def update(self, data):
      self._md5.update(data)

   def digest(self):
      return self._md5.digest()

   def __repr__(self):
      str = self.digest()
      return "%02x"*len(str) % tuple(map(ord, str))
         # This nice was suggested by Guido


   def md5file(self, f):
      if type(f) == type(''): # If f is string - use it as file's name
         infile = open(f, 'r')
      else:
         infile = f           # else assume it is opened file (fileobject) or
                              # "compatible" object (must has readline() method)

      try:
         while 1:
            buf = infile.read(16*1024)
            if not buf: break
            self.update(buf)

      finally:
         if type(f) == type(''): # If f was opened - close it
            infile.close()


if __name__ == "__main__":
   print "This must print exactly the string"
   print "Test: 900150983cd24fb0d6963f7d28e17f72"
   print "Test:", md5wrapper("abc")

From cgw at fnal.gov  Fri Nov  5 07:38:59 1999
From: cgw at fnal.gov (Charles G Waldman)
Date: Fri, 5 Nov 1999 13:38:59 +0100 (MET)
Subject: [OFFTOPIC] Maps and functions (was Re: Apply a function to each list member?)
In-Reply-To: <19991105215504.A497@Ridcully.home>
References: <slrn824gee.1f7.bwinton@tor.dhs.org>
	<XFMail.991105110617.mikael@isy.liu.se>
	<19991105215504.A497@Ridcully.home>
Message-ID: <14370.53219.667492.946075@gargle.gargle.HOWL>

Malcolm Tredinnick writes:
 > 
 > Correct. In fact, since we're being precise, let me point out that
 > (mathematically speaking), all functions are maps (or mappings), but not
 > all mappings are functions. The point is that a function should have only a
 > single "output" for any given input (i.e. it should be many-to-one or
 > one-to-one, not one-to-many). So some mapping that sends, say, the number 2
 > to the numbers 3 and 4 is not a function (but a mapping that sends both the
 > numbers 3 and 4 to the number 2 is a function, since for each input there
 > is only a single output).

Hmm... what kind of "map" are you thinking of that could send 2 to
"both 3 and 4"?  I think such a thing is not a mapping, at least not
according to any definition of "map" that I've ever seen.  If you
look, for instance, in any basic topology text, you will see that
"map" and "function" mean exactly the same thing, when used as nouns.
"Map" also has a sense as a transitive verb, which means "to apply a
function to", which is where the Python/functional programming usage
comes from.




From wahern at jinsa.org  Fri Nov  5 17:26:52 1999
From: wahern at jinsa.org (William Ahern)
Date: Fri, 5 Nov 1999 17:26:52 -0500
Subject: making the Python case to my dept.
Message-ID: <9911051740130D.02530@william.jinsa.org>

Hello,

I have never used Python, but I'm currently attempting to persuade a department
at my university (Geroge Washington) to use Python as an intro language. I have
read that it was good for this. From what I hear, the current options on the
table are Delphi and Java. I don't know much about Delphi, but I doubt Java is
any good for intro classes.

Could anybody suggest some substantive info to make the case (or not make it).
Also, the department will need an IDE (Win95/Mac), so if anybody knows where to
find one....

I know Python is being used on a broad range of applications. Complaints from
students these days, a professor told me, is that the current language, Pascal,
is outdated and doesn't have much applicability in the real world. I tend to
agree, though I'm an International Affairs major, not CS.

I've seen all of the energy and spirit behind the Python community, and I can't
fathom why any cash strapped organization would drop a few grand for something
that is freely available on the internet. Anyhow, how much support coukd
somebody need for an IDE (and a prgramming language). Not to mention the fact
that a free and OSS language implementation and IDE would preclude students
from secretly passing around the only CD in class w/ the applications on it. I
figure using Python would be a start to getting people to think outside of the
Microsoft culture (for better or worse).

In any event, please at least CC: responses to wahern at jinsa.org. I'm not
subscribed to this list.

Thanx a bunch,

Bill Ahern
MIS, JINSA



From jhefferon at my-deja.com  Fri Nov  5 16:19:06 1999
From: jhefferon at my-deja.com (jhefferon at my-deja.com)
Date: Fri, 05 Nov 1999 21:19:06 GMT
Subject: Python and Literate Programming
References: <Pine.LNX.3.96.991104011425.943A-100000@dhcp9545112> <c75iu3il9hn.fsf@pc142.cosc.canterbury.ac.nz>
Message-ID: <7vvhk7$9h4$1@nnrp1.deja.com>

  "Tim Evans" <tre17 at student.canterbury.ac.nz> wrote:
> For just this functionality, you could try using the `noweb' tool.  It
> manages to not break pythons indentation much at all.  For example,
> the input:
>
> <<example outer block>>=
> for p in parrots:
>     <<complain if the parrot p is dead>>
>
> <<check if the parrot p is alive>>=
> if p.perched and not p.nailed:
>     print "It's f****** stuffed it!"
>
> Will actually be correctly turned into:
>
> for p in parrots:
>     if p.perched and not p.nailed:
>         print "It's f****** stuffed it!"
>
> I'm not sure if other tools will do this correctly.  Noweb has the
> disadvantages that it's written in icon, and that its LaTeX output is
> not great, but it isn't all bad.
Noweb works great; that's what I use.  But looking on deja.com on the
comp.lang.programming.literate group for Python also nets a nice program
called py2tex by Jeroen van Maanen <jeroenm at cwi.nl>.  Very nice output.

Jim Hefferon


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



From aa8vb at yahoo.com  Mon Nov  1 10:21:03 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Mon, 1 Nov 1999 10:21:03 -0500
Subject: Tkinter canvas blow-up:  bbox(ALL) == None
Message-ID: <19991101102103.A134035786@vislab.epa.gov>

     Apparently if the bounds of the items in a Tk canvas widget exceed
MAXINT, bbox(ALL) returns None.  The attached Python script demonstrates
this.

     Intuitively it seems like this is a bug since the contents of the
canvas are maintained in floating point; integer bounds shouldn't be
involved, should they?

     Is this a Tk bug or a Tkinter bug?

Thanks,

Randall


-- 
Randall Hopper
aa8vb at yahoo.com
-------------- next part --------------
#!/usr/bin/env python
#
#   canvas_test.py - Demonstrates a bug in Tkinter's Canvas widget
#                    where bbox = None when the bbox exceeds signed MAXINT
#                    (2^31 on most machines).

from Tkinter import *

#
# Create widgets
#
root = Tk()
canvas = Canvas( root, width = 300, height = 300 )
canvas.pack()
canvas.create_rectangle( (-100,-100,100,100),
                         outline = "#008000", fill="#800000", width = 3 )
canvas.create_line( (-100,-100,100,100),
                    fill="blue", width=3 )
canvas.create_line( (-100,100,100,-100),
                    fill="blue", width=3 )
canvas.config( scrollregion = canvas.bbox( ALL ) )

#
# Zoom the canvas until it's bounds get toasted because they exceed MAXINT
#
def TimerCB( canvas=canvas ):
  old = canvas.bbox( ALL )
  canvas.scale( ALL, 0.0, 0.0, 2.0, 2.0 )
  new = canvas.bbox( ALL )
  print "Old = %s, New = %s" % ( old, new )
  if new == None:
    raise SystemError, "canvas's bbox is toast (exceeds signed MAXINT)"
  canvas.after( 300, TimerCB )

print "CANVAS ZOOM BLOW-UP TEST:\n"
canvas.after( 1500, TimerCB )
root.mainloop()

From a.eyre at optichrome.com  Wed Nov 24 04:40:43 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Wed, 24 Nov 1999 09:40:43 -0000
Subject: Minimal Python installation?
In-Reply-To: <MPG.129cf032cc39a31c989686@news.ericsson.se>
Message-ID: <002b01bf365f$fa007430$3acbd9c2@peridot.optichrome.com>

> We intend to bundle Python with a product, but would like to avoid
> wasting more space than necessary. Is this a problem, license wise?
> I mean, is one supposed to pass the _whole_ Python distribution only?

You can, but (assuming Win32) you only need the python15.dll, and the
contents of the Lib directory.

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From smigowsky at dspace.de  Fri Nov 12 02:56:08 1999
From: smigowsky at dspace.de (Stefan Migowsky)
Date: Fri, 12 Nov 1999 08:56:08 +0100
Subject: Win32api.... can it create a Windows shortcut (.lnk) file?
Message-ID: <326536345498D311B3BC00105A39802A0744C8@newsintern.dspace.de>

Hi, 

since this program is so clean and simple there is no error checking.
The IPersists interface just throws an unspecified OLE exception, if
i.e. the path of the lnk is non existent. So just include some error 
checking like :

...
import os
import types

if type(Path) != types.StringType or\
   type(Target) != types.StringType or\
   type(Arguments) != types.StringType or\
	....
	raise TypeError, "String arguments required"


LinkDir   = os.path.dirname(Path)
TargetDir = os.path.dirname(Target)

if not os.path.isfile(Target) or\
	raise TypeError,  "Target file does not exist"  # Not very correct
but convienient

if not os.path.isdir(LinkDir)
	raise TypeError,  "Link directory does not exist"  # Not very
correct but convienient

....  etc.


Stefan

Dipl.-Physiker Stefan Migowsky	
dSPACE GmbH		
Technologiepark 25			33100 Paderborn (Germany)
mailto:smigowsky at dspace.de 		http://www.dspace.de



-----Original Message-----
From: me [mailto:none at fake.net]
Sent: Friday, November 12, 1999 6:18 AM
To: smigowsky at dspace.de
Subject: Re: Win32api.... can it create a Windows shortcut (.lnk) file?


That's a nice clean example.... but I'm having one small problem, even
taking it exactly as you put it:


##########################################
Traceback (innermost last):
  File "C:\Program
Files\Python\Win32\Pythonwin\pywin\framework\scriptutils.py", 
line 237, in
RunScript
    exec codeObject in __main__.__dict__
  File "C:\PROGRA~1\PYTHON\MYPROJ~1\WINTOOLS\wintools.py", line 
144, in ?
    CreateShortCut(Path,Target,Arguments,StartIn,Icon,Description)
  File "C:\PROGRA~1\PYTHON\MYPROJ~1\WINTOOLS\wintools.py", line 132, in
CreateShortCut
    persist.Save(Path, 1)
com_error: (-2147467259, 'Unspecified error', None, None)
##########################################

Unfortunately, "Unspecified error" doesn't help me too much... 
q:].  I'll be
looking into that tonite, but if you have a quick fix, I'd 
appreciate it.
I've been using the win32api module a lot, and it's great, but 
not pythoncom
a lot yet...

q:]

Kevin.


Stefan Migowsky <smigowsky at dspace.de> wrote in message
news:137053459BA7D211B3930090272A62CA8EDD80 at newsintern.dspace.de...
> Hi
>
>
> Mark Hammond did a great job on wrapping the win32api. So
> creating a shortcut is just so easy :
> ################ CreateShortCut.py ####################
>
> from win32com.shell import shell
> import win32api
> import pythoncom
> import os
> import sys
>
> def CreateShortCut(Path, Target,Arguments = "", StartIn = "", Icon =
("",0),
> Description = ""):
>     # Get the shell interface.
>     sh = pythoncom.CoCreateInstance(shell.CLSID_ShellLink, None, \
>         pythoncom.CLSCTX_INPROC_SERVER, shell.IID_IShellLink)
>
>     # Get an IPersist interface
>     persist = sh.QueryInterface(pythoncom.IID_IPersistFile)
>
>     # Set the data
>     sh.SetPath(Target)
>     sh.SetDescription(Description)
>     sh.SetArguments(Arguments)
>     sh.SetWorkingDirectory(StartIn)
>     sh.SetIconLocation(Icon[0],Icon[1])
>
>     # Save the link itself.
>     persist.Save(Path, 1)
>
> if __name__ == "__main__":
> TempDir = os.environ["TEMP"]
> WinRoot = os.environ["windir"]
>
> Path        =  WinRoot    + "\\Profiles\\All Users\\Desktop\\New
> Link.lnk"
> Target      =  Pythonroot + "pythonw.exe "
> Arguments   =  TempDir + "\\test.py"
> StartIn     =  TempDir
> Icon        = (Pythonroot + "\\py.ico", 0)
> Description = "New Link"
>
> CreateShortCut(Path,Target,Arguments,StartIn,Icon,Description)
>
>
> Stefan
>
>
>
> Dipl.-Physiker Stefan Migowsky
> dSPACE GmbH
> Technologiepark 25 33100 Paderborn (Germany)
> mailto:smigowsky at dspace.de http://www.dspace.de
>
>
>
> -----Original Message-----
> From: me [mailto:none at fake.net]
> Sent: Thursday, November 11, 1999 7:54 AM
> To: python-list at python.org
> Subject: Win32api.... can it create a Windows shortcut (.lnk) file?
>
>
> I'm just trying to find an easy to create shortcuts on 
Windows platforms
> through Python... it doesn't seem to be a simple matter though.
>  I've gone
> through all the Win32 docs, and found nothing, and have 
started writing
> (with limited success) a C function to do it.
>
> Is there an easy way?  q:]
>
> Thanks,
> Kevin Cazabon
> 
>
>
>
> --
> http://www.python.org/mailman/listinfo/python-list
>
>





From mhammond at skippinet.com.au  Tue Nov 16 00:50:00 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Tue, 16 Nov 1999 05:50:00 GMT
Subject: Bug? - win32pipe.popen2 is moody under W98
References: <80qkbv$2m5$1@nntp5.atl.mindspring.net>
Message-ID: <cg6Y3.5073$Xo.26549@news-server.bigpond.net.au>

Eric Johnson wrote in message <80qkbv$2m5$1 at nntp5.atl.mindspring.net>...

>Can anyone say either way whether this can be expected to work under
>W98?

Unfortunately it is not.  This is mentioned in the Python FAQ (including a
link to the MS KB article that documents the bug)

On the brighter side, Bill Tutt finally decided to work around this bug in
Windows 98.  It will be there in the next version of win32all.  In the
mean-time, I could probably send you the necessary updates...  However, it
will be pretty resource hungy - the solution (as documented by MS) is to
execute a "shim" exe - so each popen on Win95/98 will now have _3_ new
processes - this shim, command.com and the actual .exe you are trying to
run.  Win95/98 probably wont handle the strain too well, thereby giving you
a practical limit of just a few popen's running...

Mark.






From gerrit.holl at pobox.com  Tue Nov 30 08:12:24 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Tue, 30 Nov 1999 14:12:24 +0100
Subject: Htmlizing text
In-Reply-To: <19991129170705.A314@better.net>; from parkw@better.net on Mon, Nov 29, 1999 at 05:07:05PM -0500
References: <943909426snz@vision25.demon.co.uk> <19991129170705.A314@better.net>
Message-ID: <19991130141224.A2813@stopcontact.palga.uucp>

William Park wrote:
> On Mon, Nov 29, 1999 at 09:03:46PM +0000, Phil Hunt wrote:
> > 
> > Is there a function in the standard Python library to HTML-ize text,
> > i.e. to replace 'a > b & c < d' with 'a &gt; b &amp; c &lt; d'?
> > 
> > (I realize this can be done with regular expressions, but is there
> > an easy way?)
> > 
> > -- 
> > Phil Hunt - - - phil at comuno.com
> 
> The following function replaces [<&>] unconditionally.
>     def escapeall(s):
> 	s = string.replace(s, '&', '&amp;')		# do this first
> 	s = string.replace(s, '<', '&lt;')
> 	s = string.replace(s, '>', '&gt;')
> 	return s
> I used this in my 'HTMLtag' module which generates HTML tags.

This is better:

def escapeall(s):
	import htmlentitydefs
	for k, v in htmlentitydefs.items():
		s = string.replace(s, v, '&' + k)
	return s

By the way, why doesn't the cgi module does this?

regards,
Gerrit.

-- 
"The move was on to 'Free the Lizard'"

  -- Jim Hamerly and Tom Paquin (Open Sources, 1999 O'Reilly and Associates)
  2:08pm  up 35 min,  4 users,  load average: 0.00, 0.00, 0.08



From mhammond at skippinet.com.au  Sun Nov  7 00:20:48 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sun, 07 Nov 1999 05:20:48 GMT
Subject: win32com and IIS?
References: <802s6a$mj5$1@web1.cup.hp.com>
Message-ID: <Q_7V3.5012$pf4.18247@news-server.bigpond.net.au>

I have never seen this before.  However, these details really arent enough
to help me diagnose the problem.  It is possible to break this down further?

Mail me if you want help doing this.  A good start would be to turn on the
debugging flags at the top of win32com\client\dynamic.py...

Mark.

Arad Rostampour wrote in message <802s6a$mj5$1 at web1.cup.hp.com>...
>Hi,
>
>I've been using win32com to access IE and get webpages using:
>
>import win32com.client
>ie = win32com.client.Dispatch("InternetExplorer.Application")
>ie.Navigate("http://www.hp.com")
>...
>
>This works fine from the command line.  But, when I try to do this from a
>CGI script under IIS, I get the following error:
>
>File "f:\python\win32com\client\dynamic.py", line 319, in __getattr__ raise
>pythoncom.com_error, details pywintypes.com_error: (-2147467259,
>'Unspecified error', (0, None, None, None, 0, -2147467259), None) Exception
>exceptions.AttributeError: in ignored
>
>Any ideas on how to work around this?






From stidolph at origin.ea.com  Tue Nov 23 15:23:10 1999
From: stidolph at origin.ea.com (Stidolph, David)
Date: Tue, 23 Nov 1999 14:23:10 -0600
Subject: Launching Python debugger from embedded Python
Message-ID: <11A17AA2B9EAD111BCEA00A0C9B41793034AAF19@molach.origin.ea.com>

I'm getting my company to register for PyWin so we can get the debug
versions of PyWin, but in the meantime I made a release version of our app
to test it out.  I have the following code that is called from a Menu
operation in our own ui:

  def OnMenuDebug(self,x,y,flags):
    print 'OnMenuDebug'
    import pywin.debugger
    try:
      a = 1 / 0
    except:
      pywin.debugger.post_mortem()

In the Output Debug Window of MSDev, I can see that the win32ui.pyd is
loading, and it loads other DLL's, but it crashes.  I'm going to do another
release build with all the debug info turned on, but wanted to get this
question out to see if anyone has ideas to get the debugger working.

OnMenuDebug
Loaded 'C:\Program Files\Python\Pythonwin\win32ui.pyd', no matching symbolic
information found.
Loaded 'C:\WINNT\system32\PyWinTypes15.dll', no matching symbolic
information found.
Loaded 'C:\WINNT\system32\ole32.dll', no matching symbolic information
found.
Loaded 'C:\WINNT\system32\oleaut32.dll', no matching symbolic information
found.
Loaded 'C:\WINNT\system32\mfc42.dll', no matching symbolic information
found.
Loaded 'C:\WINNT\system32\winspool.drv', no matching symbolic information
found.
Loaded 'C:\WINNT\system32\shell32.dll', no matching symbolic information
found.
Loaded 'C:\WINNT\system32\shlwapi.dll', no matching symbolic information
found.
First-chance exception in uo2client.exe (PYTHON15.DLL): 0xC0000005: Access
Violation.
The program 'C:\UO2\build-win32\uo2client\Release\uo2client.exe' has exited
with code 0 (0x0).





From A.Weis at t-online.de  Sat Nov 27 18:05:58 1999
From: A.Weis at t-online.de (Arnold Weis)
Date: Sat, 27 Nov 1999 23:05:58 GMT
Subject: Sending files to a WWW-Server
References: <81nje2$su9$1@rznews.rrze.uni-erlangen.de>
Message-ID: <38406354.3417747@news.main-echo.net>

"Jan Stiller" <stiller at phil.uni-erlangen.de> wrote:

>Hi all
>
>I'm searching for examples how to send files to a web server. I've found
>out, that I have to use multipart messages, but don't know exactly how to
>construct them.

Maybe http://starship.python.net/crew/jbauer/cgiupload/ is what you
are searching for?

Regards

	Arnold



From ionel at psy.uva.nl  Tue Nov 16 09:55:11 1999
From: ionel at psy.uva.nl (Ionel Simionescu)
Date: Tue, 16 Nov 1999 15:55:11 +0100
Subject: docs
References: <000a01bf2ffc$3e70b0e0$042d153f@tim>
Message-ID: <80rr5j$fku@mail.psy.uva.nl>

Tim Peters <tim_one at email.msn.com> wrote in message
news:000a01bf2ffc$3e70b0e0$042d153f at tim...
| Ask MS for the source code <wink>.  Short of that, there's no reason to
| suppose it's a problem with the docs, and every reason to suspect the
| obvious.

Indeed, there was no problem with the docs.
Our guess was that the IE would register properly the MIME type (for the
cached copies), only if the comment between <HTML> and <BODY> is moved
elsewhere.

ionel





From ivanlan at callware.com  Tue Nov 30 22:00:49 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Tue, 30 Nov 1999 20:00:49 -0700
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <81bmmi$279$1@nnrp1.deja.com> <3841EE77.6B5D6AC9@wjk.mv.com> <FLxz96.IEI@world.std.com> <38435C43.B238EC15@mindspring.com> <FM0KtD.2ps@world.std.com> <m33dtoro6a.fsf@atrus.jesus.cam.ac.uk> <384484A2.AF4BE448@mindspring.com>
Message-ID: <38448F61.6153AC89@callware.com>

Hi All---

"Jesse D. Sightler" wrote:

[horrible stuff bobbitted]

> > >>> incr(x)
> > >>> x
> > 2
> 
> Ack, and I though Perl was the only language that made you do stuff like
> that.  :)  Could you please explain this little piece of code?
> 
> It sure looks like byte-code dependant, self-modifying crazyness to me.
> :)
> 

NO, no, don't ask him to explain it!  It'll just encourage him!

<ain't-no-prejudices-here>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours
----------------------------------------------



From news at dorb.com  Tue Nov 30 16:42:06 1999
From: news at dorb.com (Darrell)
Date: Tue, 30 Nov 1999 16:42:06 -0500
Subject: Does this code create a circular reference ?
References: <8216ih$2ki@mail.psy.uva.nl>
Message-ID: <XuX04.11611$wG3.742951@typhoon.nyroc.rr.com>

Let the child keep a key to look up the parent.

dic={}

 class child:
   def __init__(self, parent):
     self.parent = parent

   def do_something(self):
     pass

 class headache:
    def __init__(self):
        global dic
        dic[`self`]=self

   def method(self):
     a_child = child(`self`)
     a_child.do_something()


--
--Darrell
"Ionel Simionescu" <ionel at psy.uva.nl> wrote in message > I need advice on
whether the situation described below introduces circular
> references.
>






From wlfraed at ix.netcom.com  Wed Nov 24 11:26:01 1999
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Wed, 24 Nov 1999 08:26:01 -0800
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <8666ysfta9.fsf@ix.netcom.com> <012e01bf365d$912fa360$f29b12c2@secret.pythonware.com>
Message-ID: <5g08OLe6ncyc5Uu5627KDJVZ2N0R@4ax.com>

On Wed, 24 Nov 1999 10:23:27 +0100, "Fredrik Lundh"
<fredrik at pythonware.com> declaimed the following in comp.lang.python:

> Brad Knotwell <knotwell at ix.netcom.com> wrote:
> > Similarly, the only thing I've found missing from Python is C's
> > ternary operator, it would be nice to be able to do something like 
> > the following:
> > 
> > def max(x,y): return ((x>y) ? x : y)
> 
> now that was one lousy example ;-)
>
	Also potentially unsafe if either x or y is some expression when
invoked. At least, I seem to recall some suggestion to always put macro
arguments in parens...

	def max(x,y): return(((x)>(y)) ? (x) : (y))

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



From gmcm at hypernet.com  Sun Nov 14 22:24:12 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sun, 14 Nov 1999 22:24:12 -0500
Subject: Variable scope question.. singletons?
In-Reply-To: <382f70a7@news.nwlink.com>
Message-ID: <1269486246-1025196@hypernet.com>

Adam Vandenberg asks:

> I have a bunch of modules that will be used by a bunch of top-level scripts
> (cgis).
> Each top-level script is going to: "import prefs" which I want to check
> cookies
> and fill out a Prefs class.
> So far OK.
> 
> Now, is there any way to make this class "more global" so that any of the
> modules I import can access it?
> I guess this is sort of like declaring the variable "extern".
> 
> Any thoughts? 

Sometimes. Rarely on Sundays.

> Does this make sense?

Not an iota of a scintilla <wink>.

The very first import will install the module in sys.modules. 
Subsequent imports will simply return references to the thing 
in sys.modules. So, in the absense of bizarre import hooks, it 
already is "extern".

returning-to-Sunday-night-oblivion-ly y'rs

- Gordon



From sekter at mail.matav.hu  Mon Nov 15 09:51:35 1999
From: sekter at mail.matav.hu (Arpad Kiss)
Date: Mon, 15 Nov 1999 15:51:35 +0100
Subject: PIL 1.0 on win32?
References: <009b01bf2f5e$fef3b900$029f38c3@geometria.hu> <01ee01bf2f61$96097a80$f29b12c2@secret.pythonware.com> <004701bf2f6a$bd67f990$029f38c3@geometria.hu>
Message-ID: <005501bf2f78$ea3d9a70$029f38c3@geometria.hu>

Hi,
I've got some new font files and with them it works.
And when I commented out the text drawing part of my code I left some text
related line in it, that causes the mirrored image(i think).
Thanks a lot,
Arpad Kiss

----- Original Message -----
From: Arpad Kiss <sekter at mail.matav.hu>
To: Fredrik Lundh <fredrik at pythonware.com>
Cc: <python-list at python.org>
Sent: Monday, November 15, 1999 2:10 PM
Subject: Re: PIL 1.0 on win32?


>
> Fredrik Lundh wrote:
> > Arpad Kiss wrote:
> > > I have a PIL based Zope product I have been developing on a Linux box.
I
> > > just wanted to test it on a Windows NT(sp5), but it doesn't work.:-(
> > > At first it didn't recognize my fonts, because the readline method on
> win32
> > > returns 10+13 at the end of the lines and in ImageFont.py only \n is
> > > checked.
> >
> > PIL opens the font files in binary mode, so this
> > shouldn't be a problem.  you probably messed
> > something up when you transferred the font
> > files to the PC...
>
> I have created a .tar.gz then extracted it with WinZip.
>
> >
> > > After I  solved it I got an error message(memory error in line 151 of
> > > ImageDraw.py) when I wanted to draw some text.
> > > When I commented out the text drawing code in my program, then I got
> back a
> > > picture, but it is mirrored!
> >
> > this sounds really weird -- but at least the
> > memory error might be due to bogus font data...
>
> Ok. I will try to replace my fonts.
>
> >
> > what win32 build do you use?  can you post some
> > code?
>
> Python1.5.2 and pil-win32-991101.zip
> I will write a minimal example.
>
> >
> > > Of course on linux box my program works fine.
> > > I just want know who else use PIL 1.0 on win32 and what experiences
they
> > > have. Should this text drawing function work? Is it my fault?
> >
> > fwiw, PIL is being developed on win32...
>
> Maybe a minimal example(without Zope) can help me, if cannot then I'll
post
> it.
> Thanks,
> Arpad
>
>
>
> --
> http://www.python.org/mailman/listinfo/python-list
>




From fredrik at pythonware.com  Tue Nov  2 05:22:23 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 2 Nov 1999 11:22:23 +0100
Subject: Comparison between Python and "Ruby"
References: <1270663667-24391282@hypernet.com><381ED7A6.D7D7FB13@compaq.com> <877lk1wiij.fsf@ev.netlab.co.jp>
Message-ID: <003901bf251c$290775d0$f29b12c2@secret.pythonware.com>

Yukihiro Matsumoto <matz at netlab.co.jp> wrote:
> I guess he loves reducing variablity, if possible, which is totally
> against TMTOWTDI.  I don't judge whether it is good or bad.

you're confusing syntax with semantics.

there are MANY ways to solve any given problem
in Python, but when you've decided that you want
to use a certain pattern (be it a loop, an iterator,
an object factory, or any other pydiom) in your
solution, there's usually ONE way to write the
code.

think about it.

</F>

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




From hildeb at www.stahl.bau.tu-bs.de  Sun Nov 28 10:47:16 1999
From: hildeb at www.stahl.bau.tu-bs.de (Ralf Hildebrandt)
Date: 28 Nov 1999 15:47:16 GMT
Subject: fcntl/ioctl and O_NONBLOCK
References: <slrn8405je.q0.hildeb@stahlw06.stahl.bau.tu-bs.de> <slrn840205.sn4.hildeb@stahlw06.stahl.bau.tu-bs.de> <slrn83t5mv.bhk.hildeb@stahlw06.stahl.bau.tu-bs.de> <81ml5t$eer$0@216.39.151.169> <81p32a$m14$1@216.39.151.169> <81qj26$be6$0@216.39.151.169>
Message-ID: <slrn842jk3.o62.hildeb@stahlw06.stahl.bau.tu-bs.de>

On 28 Nov 1999 06:45:26 GMT, Donn Cave <donn at oz.net> wrote:

>Yes - only a guess, but it bites everyone at least once when trying
>to talk to another process via pipes.  After each write(), where your
>Python program is writing to the ``stdin'' of the plot process, add
>a flush(), i.e. stdin.flush().

Yup, forgot that one. As usual.

>data to be written immediately anyway.  You can also control this
>by the buffer size parameter to the function that creates the file
>object, or just avoid a file object altogether and write directly
>to the pipe file descriptor.
Yes? I always wondered what the "default" buffer size was.



From Bjoern.Giesler at stud.uni-karlsruhe.de  Tue Nov 23 13:11:40 1999
From: Bjoern.Giesler at stud.uni-karlsruhe.de (Bjoern.Giesler at stud.uni-karlsruhe.de)
Date: Tue, 23 Nov 1999 19:11:40 +0100 (CET)
Subject: Embedding: Getting the module from a PyObject *
Message-ID: <E11qKPm-0004bF-00@rzstud1.rz.uni-karlsruhe.de>

Hi,

can anyone tell me how to get the module object that a certain instance's
class is defined in, from that instance? 

In my code, I call PyImport_ImportModule(name). The module's Python code 
defines a plugin object obj and calls CPlugin.register(obj), which maps 
to the C function CPluginPyRegister().

In CPluginPyRegister, I want to add the module's PyObject * to an
internal structure so I can unload the module at some later time.

Phew, I hope this isn't too unclear. If anyone could help me, I'd be
extremely thankful.

		--Bjoern



From mikael at isy.liu.se  Fri Nov  5 07:32:57 1999
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Fri, 05 Nov 1999 14:32:57 +0200 (MET)
Subject: [OFFTOPIC] Maps and functions (was Re: Apply a function to e
In-Reply-To: <14370.53219.667492.946075@gargle.gargle.HOWL>
Message-ID: <XFMail.991105143257.mikael@isy.liu.se>

On 05-Nov-99 Charles G Waldman wrote:
 >  Hmm... what kind of "map" are you thinking of that could send 2 to
 >  "both 3 and 4"?  I think such a thing is not a mapping, at least not
 >  according to any definition of "map" that I've ever seen.  If you
 >  look, for instance, in any basic topology text, you will see that
 >  "map" and "function" mean exactly the same thing, when used as nouns.
 >  "Map" also has a sense as a transitive verb, which means "to apply a
 >  function to", which is where the Python/functional programming usage
 >  comes from.

There is nothing that says that a map cannot output a set. But then again, 
that is an object (in a set of sets), even if the set consists of many 
smaller objects. As an example, the inverse mapping of the squaring 
function (a->a^2) is a mapping from the set of reals to a set of sets of 
reals, sets containing at most two real elements (b->{+/-sqrt(b)},{0}, or 
the empty set), depending on if b is positive, zero, or negative.

You are probably right when you state that a function is a mapping is
a function, but people tend to give them different interpretations.
As an example (to make things even more confused), I have a reference 
(Davenport: Probability and Random Processes) that uses the words in the 
following way: 

  A function determines a mapping from one set to another. 

This suggests that a function is just an expression, and the mapping is 
its action on the elements of a set.

Are-we-being-too-scientific-ly y'rs

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    05-Nov-99
Time:    13:55:12

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



From fredrik at pythonware.com  Wed Nov 24 06:50:45 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 24 Nov 1999 12:50:45 +0100
Subject: Absolute newbie question
References: <Pine.SOL.3.96.991124165729.11616B-100000@giasbma>
Message-ID: <002501bf3672$26107e50$f29b12c2@secret.pythonware.com>

Philip Tellis <tellis at giasbma.vsnl.net.in> wrote:
> Where are the man pages for python?  I am new to python, and would like to
> look through the man pages before I start.  Someone please help.

$ man python

NAME
       python - an interpreted, interactive, object-oriented pro-
       gramming language

SYNOPSIS
       python [ -d ] [ -i ] [ -O ] [ -S ] [ -t ] [ -u ] [ -v ]  [
       -x ] [ -X ]
              [ -c command | script | - ] [ arguments ]

DESCRIPTION
       Python  is  an  interpreted,  interactive, object-oriented
       programming language that combines remarkable  power  with
       very  clear syntax.  For an introduction to programming in
       Python you are  referred  to  the  Python  Tutorial.   The
       Python  Library  Reference documents built-in and standard
       types, constants, functions  and  modules.   Finally,  the
       Python Reference Manual describes the syntax and semantics
       of the core language in (perhaps too) much detail.

       Python's basic power can be extended with your own modules
       written  in C or C++.  On most systems such modules may be
       dynamically loaded.  Python is also adaptable as an exten-
       sion language for existing applications.  See the internal
       documentation for hints.

COMMAND LINE OPTIONS
       -d     Turn on parser debugging output (for wizards  only,
              depending on compilation options).

       -i     When a script is passed as first argument or the -c
              option is used, enter interactive mode  after  exe-
              cuting the script or the command.  It does not read
              the $PYTHONSTARTUP file.  This  can  be  useful  to
              inspect  global  variables  or a stack trace when a
              script raises an exception.

       -O     Turn on  basic  optimizations.   This  changes  the
              filename  extension  for  compiled (bytecode) files
              from .pyc to pyo.

       -S     Disable the import of the module site and the site-
              dependent   manipulations   of   sys.path  that  it
              entails.

       -t     Issue a warning when a source file mixes  tabs  and
              spaces  for  indentation  in  a  way  that makes it
              depend on the worth of a tab expressed  in  spaces.
              Issue an error when the option is given twice.

       -u     Force  stdin,  stdout  and  stderr  to  be  totally
              unbuffered.

       -v     Print a message each time a module is  initialized,
              showing  the  place  (filename  or built-in module)


       -x     Skip  the  first  line  of  the  source.   This  is
              intended  for  a  DOS specific hack only.  Warning:
              the line numbers in error messages will be  off  by
              one!

       -X     Make  the  standard  exceptions  strings instead of
              classes.  Use for backward compatibility  with  old
              code only.

       -c command
              Specify  the command to execute (see next section).
              This terminates the option list (following  options
              are passed as arguments to the command).

INTERPRETER INTERFACE
       The  interpreter  interface  resembles  that  of  the UNIX
       shell: when called with standard input connected to a  tty
       device, it prompts for commands and executes them until an
       EOF is read; when called with a file name argument or with
       a  file  as standard input, it reads and executes a script
       from that file; when called with -c command,  it  executes
       the  Python  statement(s)  given as command.  Here command
       may contain multiple  statements  separated  by  newlines.
       Leading  whitespace  is  significant in Python statements!
       In  non-interactive  mode,  the  entire  input  is  parsed
       befored it is executed.

       If  available,  the  script  name and additional arguments
       thereafter are passed to the script in the Python variable
       sys.argv  ,  which  is  a  list of strings (you must first
       import sys to be able to access it).  If no script name is
       given,  sys.argv is empty; if -c is used, sys.argv[0] con-
       tains the string '-c'.  Note that options  interpreted  by
       the  Python interpreter itself are not placed in sys.argv.

       In interactive mode, the primary prompt is `>>>'; the sec-
       ond  prompt (which appears when a command is not complete)
       is `...'.  The prompts can be  changed  by  assignment  to
       sys.ps1  or  sys.ps2.  The interpreter quits when it reads
       an EOF at a prompt.  When an unhandled exception occurs, a
       stack  trace is printed and control returns to the primary
       prompt; in non-interactive  mode,  the  interpreter  exits
       after  printing  the  stack  trace.   The interrupt signal
       raises the KeyboardInterrupt exception; other UNIX signals
       are  not caught (except that SIGPIPE is sometimes ignored,
       in favor of the IOError exception).   Error  messages  are
       written to stderr.

FILES AND DIRECTORIES
       These are subject to difference depending on local instal-
       lation conventions:

       /usr/local/bin/python
              Recommended location of the interpreter.

       /usr/local/lib/python<version>
              Recommended location of  the  directory  containing
              the standard modules.

ENVIRONMENT VARIABLES
       PYTHONHOME
              Change   the   location   of  the  standard  Python
              libraries.  By default, the libraries are  searched
              in   <prefix>/lib/python<version>   and  <exec_pre-
              fix>/lib/python<version>,   where   <prefix>    and
              <exec_prefix>  are  installation-dependent directo-
              ries, both defaulting to /usr/local.  When $PYTHON-
              HOME  is  set  to  a  single  directory,  its value
              replaces both <prefix> and <exec_prefix>.  To spec-
              ify  different values for these, set $PYTHONHOME to
              <prefix>:<exec_prefix>.

       PYTHONPATH
              Augments the default search path for module  files.
              The format is the same as the shell's $PATH: one or
              more directory pathnames separated by colons.  Non-
              existant  directories  are  silently  ignored.  The
              default search path is installation dependent,  but
              generally  begins with <prefix>/lib/python<version>
              (see PYTHONHOME below).  The default search path is
              always  appended to $PYTHONPATH.  If a script argu-
              ment is given, the directory containing the  script
              is  inserted  in  the path in front of $PYTHONPATH.
              The search path can be manipulated  from  within  a
              Python program as the variable sys.path .

       PYTHONSTARTUP
              If  this is the name of a readable file, the Python
              commands in that file are executed before the first
              prompt  is displayed in interactive mode.  The file
              is executed in the same name space  where  interac-
              tive  commands are executed so that objects defined
              or imported in it can be used without qualification
              in  the  interactive  session.  You can also change
              the prompts sys.ps1 and sys.ps2 in this file.

       PYTHONDEBUG
              If this is set to a non-empty string it is  equiva-
              lent to specifying the -d option.

       PYTHONINSPECT
              If  this is set to a non-empty string it is equiva-
              lent to specifying the -i option.

       PYTHONUNBUFFERED
              If this is set to a non-empty string it is  equiva-
              lent to specifying the -u option.

       PYTHONVERBOSE
              If  this is set to a non-empty string it is equiva-
              lent to specifying the -v option.

SEE ALSO
       Python Tutorial
       Python Library Reference
       Python Reference Manual

AUTHOR
       Guido van Rossum
       CNRI
       1895 Preston White Drive
       Reston, VA 20191
       USA

       E-mail: guido at cnri.reston.va.us, guido at python.org

       And a cast of thousands.

INTERNET RESOURCES
       Web site: http://www.python.org
       FTP site: ftp://ftp.python.org
       Newsgroup: comp.lang.python

COPYRIGHT
       Copyright 1991-1995  by  Stichting  Mathematisch  Centrum,
       Amsterdam, The Netherlands.

              All Rights Reserved

       Permission to use, copy, modify, and distribute this soft-
       ware and its documentation for any purpose and without fee
       is  hereby  granted,  provided  that  the  above copyright
       notice appear in all copies and that both  that  copyright
       notice  and  this  permission  notice appear in supporting
       documentation, and that the names  of  Stichting  Mathema-
       tisch  Centrum or CWI or Corporation for National Research
       Initiatives or CNRI not be used in advertising or  public-
       ity  pertaining  to  distribution  of the software without
       specific, written prior permission.

       While CWI is the initial source for this software, a modi-
       fied  version  is  made  available  by the Corporation for
       National  Research  Initiatives  (CNRI)  at  the  Internet
       address ftp://ftp.python.org.

       STICHTING  MATHEMATISCH CENTRUM AND CNRI DISCLAIM ALL WAR-
       RANTIES  WITH  REGARD  TO  THIS  SOFTWARE,  INCLUDING  ALL
       IMPLIED  WARRANTIES  OF MERCHANTABILITY AND FITNESS, IN NO
       EVENT SHALL STICHTING  MATHEMATISCH  CENTRUM  OR  CNRI  BE
       LIABLE  FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES
       OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA
       OR  PROFITS,  WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE
       OR OTHER TORTIOUS ACTION, ARISING OUT OF OR IN  CONNECTION
       WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.

(END)




From alan.gauld at gssec.bt.co.uk  Mon Nov 29 08:36:32 1999
From: alan.gauld at gssec.bt.co.uk (Alan Gauld)
Date: Mon, 29 Nov 1999 13:36:32 +0000
Subject: NEWBIE
References: <81m7m3$82$1@news.qub.ac.uk>
Message-ID: <38428160.56B1D46D@gssec.bt.co.uk>

Stuarty at excite.co.uk wrote:
> I have just recently started using Python, after the first day I was further
> on than I am after a semester of C programming(were at arrays...).
> tutorial by Giudo van Rossum was very useful, but I attempted printing out the
> reference library..there are now less considerably less trees in the world.

So are you asking a question?
....or just boasting? ;-)

Alan g.
-- 
=================================================
This post represents the views of the author 
and does not necessarily accurately represent 
the views of BT.



From thomas at xs4all.nl  Wed Nov  3 17:18:27 1999
From: thomas at xs4all.nl (Thomas Wouters)
Date: Wed, 3 Nov 1999 23:18:27 +0100
Subject: BSDI 4.0, threads and signals.
Message-ID: <19991103231827.H29354@xs4all.nl>

I've been busy the last two days trying to track down a weird bug in BSDI
4.0.1 compiled python with threads. The problem is that the test_signal test
hangs. Compiling without thread support makes it work fine. So i whipped out
the old gdb ;)

I've tracked it down to the pause() library call. Apparently, compiling with
threads support yields very different libc code for the pause() call. And
the thread version does react to a delivered signal, the signal handler gets
called, but after that, the pause()-code does not return. It does do
something, but gcc -g for some reason doesn't link against a libc with debug
symbols, so i can't exactly see what :P

However, now i'm sort-of stuck. I could but BSDI support, but i doubt it
would bring much reaction without a test program that shows the bug -- and i
dont know how to make one :) I know the basics of threading and even posix
threads, but not the details. Might i be missing something in the way of
configuring the behaviour of pause() in a threaded enviroment ? And what
need I do to compile a teensy-tiny bit of program with threading ? I tried
the thread #defines from python's config.h et al, but that didn't work.

Anyone with any C-clue thread-wise able to give me some pointers ? It would
be greatly appreciated.

Not that this bug needs fixing for me to use python... So i'm not really in
a hurry or anything ;)

regard's'ly-y'rs,
-- 
Thomas Wouters <thomas at xs4all.net>

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



From didier.willame at ieee.org  Mon Nov  8 00:47:41 1999
From: didier.willame at ieee.org (Didier WILLAME)
Date: Mon, 08 Nov 1999 00:47:41 -0500
Subject: TEST
Message-ID: <382663FD.1C9A5B6E@ieee.org>

Sorry,

This is a test.



From sbarron at twilight.  Fri Nov  5 23:24:41 1999
From: sbarron at twilight. (Scott Barron)
Date: Sat, 06 Nov 1999 04:24:41 GMT
Subject: Writing to files from C
Message-ID: <d4OU3.3456$Gj3.180595@typ11a.deja.bcandid.com>

Hey all,

Disregard my post on this subject.  I had fflush() in the wrong function.
How silly I feel.

Scott



From mwh21 at cam.ac.uk  Mon Nov 22 08:18:19 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: Mon, 22 Nov 1999 13:18:19 +0000 (GMT)
Subject: bytecodehacks-0.60
Message-ID: <Pine.LNX.4.10.9911221304430.699-100000@localhost.localdomain>

Just a quick note to say that the (public domain) bytecodehacks have,
after a lengthy fallow period, reached version 0.60. The approach of the
version numbers to 1.0 should not be taken to mean anything.

Not much has changed; some things are neater/safer, I've corrected a few
typos in the docs and incorporated a couple of changes suggested by Evan
Simpson.

Oh yes, there's a prototype Python assembler. Most useful, I'm sure!

The package is now distributed using the distutils. Is this the first
package to do so?

Comments are, as ever, appreciated.

Regards,
Michael

<P><A HREF="http://starship.python.net/crew/mwh">bytecodehacks 0.60</A>
A minor update to the package that lets you do unlikely things in Python
by manipulating strings of bytecode. - (1999/11/22)





From skip at mojam.com  Thu Nov 18 16:50:07 1999
From: skip at mojam.com (Skip Montanaro)
Date: Thu, 18 Nov 1999 15:50:07 -0600 (CST)
Subject: urllib.getsize()
In-Reply-To: <19991118213925.A10221@optiplex.palga.uucp>
References: <19991118213925.A10221@optiplex.palga.uucp>
Message-ID: <14388.29839.294164.707388@dolphin.mojam.com>

    Gerrit> I've written a function to get the size of a HTTP/FTP url, is it
    Gerrit> good enough to be put in urllib.py?

My only worry is that HEAD doesn't always work (for whatever reason having
to do more with servers than clients).  Would it be better to wrap things in
a try/except block?  Anyone have a URL that fails with HEAD?

Skip Montanaro | http://www.mojam.com/
skip at mojam.com | http://www.musi-cal.com/
847-971-7098   | Python: Programming the way Guido indented...



From boncelet at udel.edu  Fri Nov 26 03:36:40 1999
From: boncelet at udel.edu (Charles Boncelet)
Date: Fri, 26 Nov 1999 08:36:40 +0000
Subject: DDD and Numerical Python
References: <Coe%3.8191$cz3.292560@nnrp1.uunet.ca>
Message-ID: <383E4698.E07B9FE@udel.edu>

Greg Wilson wrote:
> 
> Hi.  Has anyone integrated DDD's data visualization with the Numerical
> Python extensions?
> Thanks,
> Greg
I've used DDD a little (once or twice) and I was going to say that this
worked already (I recall seeing a plot of some NumPy data), but just now
I tried to do it and DDD crashed.

Perhaps a more appropriate request is "has anyone integrated data
visualization into IDLE"?

	Charlie Boncelet
-- 
Charles Boncelet <boncelet at udel.edu>
University of Delaware
Newark DE 19716 USA
http://www.eecis.udel.edu/~boncelet/



From news at dorb.com  Tue Nov 30 17:15:15 1999
From: news at dorb.com (Darrell)
Date: Tue, 30 Nov 1999 17:15:15 -0500
Subject: Does this code create a circular reference ?
References: <8216ih$2ki@mail.psy.uva.nl>
Message-ID: <1_X04.11614$wG3.744186@typhoon.nyroc.rr.com>

After a closer look, No it doesn't create a circular reference.
If the "headache" class had kept a reference to a_child it would have.

 def eater():
     l1=[]
     l2=[]
     l1.append(l2)
     l2.append(l1)

 for y in range(50000):
    eater()

Now that's a circular reference.

--
--Darrell
"Ionel Simionescu" <ionel at psy.uva.nl> wrote in message >

 I need advice on whether the situation described below introduces circular
> references.
>






From da at ski.org  Tue Nov  2 14:01:39 1999
From: da at ski.org (David Ascher)
Date: 2 Nov 1999 13:01:39 -0600
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Nov  2)
Message-ID: <D23AE3FFC6BD29D8.3FD0F972BD236599.A559130F1465280F@lp.airnews.net>

Today is election day in the United States.  Here are the winners from a
much simpler yet equally non-scientific selection process:

    The Python Consortium is started!  Hewlett-Packard and LLNL provides
    the marquee status and the big bucks, Digital Creations the wildly
    successful application, Interet quietly chips in, along with Foretec
    Seminars, the first organization that had a check in hand!
        http://www.python.org/consortium/

    Michel Pelletier (of Digital Creations) and Dan York (of LinuxCare)
    announce that they are writing a Zope book for O'Reilly and Associates:
        http://www.zope.org/Members/michel/ZopeBook/index_html/index_html
  
    Jan Walter announces some scripts for Blender, a computer graphics
    rendering tool, with very nice pictures to boot:
        http://www.q-bus.de/Blender/

    David S. Harrison puts forth code which formats number for human
    consumption:
        http://www.deja.com/getdoc.xp&=542270123

    Scot Hacker explains how good scripting BeOS is
	http://www.byte.com/column/BYT19991026S0001

    Tim Peters sort of sorts out some of the sort issues sorted out
    between 1.5.1 and 1.5.2:
        http://www.deja.com/getdoc.xp&=AN=541957314
  
    The same Tim accounts for a surprising 'feature' involving testing
    the object identity of immutable objects which puzzled Pearu
    Peterson and yours truly:
        http://www.python.org/pipermail/matrix-sig/1999-November/003106.html


=========================================================================
Everything you want is probably one or two clicks away in these pages:

      Python.org's Python Language Website is the center of Pythonia
          http://www.python.org

      Much of Python's real work takes place on Special-Interest Group
          mailing lists http://www.python.org/sigs/

      Python Consortium emerges as an independent nexus of activity
          http://www.python.org/consortium

      Python To-Do List anticipates some of Python's future direction
          http://www.python.org/cgi-bin/todo.py

      Python Journal is at work on its second issue
          http://www.pythonjournal.com

      Links2Go is a new semi-automated link collection; it's impressive
      what AI can generate
          http://www.links2go.com/search?search=python

    Archive probing trick of the trade:
        http://www.dejanews.com/dnquery.xp?QRY=&DBS=2&ST=PS&defaultOp=AND&LNG=ALL&format=threaded&showsort=date&maxhits=100&groups=comp.lang.python


Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://purl.org/thecliff/python/url.html
or
  http://www.dejanews.com/dnquery.xp?QRY=~g%20comp.lang.python%20Python-URL%21


Suggestions/corrections for next week's posting are always welcome.
  http://www.egroups.com/list/python-url-leads/

To receive a new issue of this posting in e-mail each Monday morning,
ask <claird at neosoft.com> to subscribe.  Mention "Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.
-- 

Cameron Laird           http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX



From ejr at CS.Berkeley.EDU  Sun Nov 28 02:46:20 1999
From: ejr at CS.Berkeley.EDU (Edward Jason Riedy)
Date: Sat, 27 Nov 1999 23:46:20 -0800
Subject: Modules/getpath.c chases links?
Message-ID: <199911280746.XAA29926@lotus.CS.Berkeley.EDU>

Is there a reason why Modules/getpath.c chases links to get the
`correct' path?  I'd like to see that removed, but the comments
make me think someone had a specific reason for including it.

I use symlink trees to separate out different versions of packages 
distributed by different people (here, stock python 1.5.2 and the
LLNL-distributed  NumPy).  The symlink chasing breaks this badly
by looking into the symlinked directory rather than using the tree
of symlinks.

Simply removing the code protected by HAVE_READLINK makes my life 
much nicer.  I'd really like to see it officially removed.  (If you
want a full-featured system for finding the installation, check the
kpathsea library in the Web2C distribution of TeX.)

Jason (not subscribed, please cc)



From Gaetan_Corneau at baan.com  Thu Nov 11 09:17:04 1999
From: Gaetan_Corneau at baan.com (Gaetan Corneau)
Date: Thu, 11 Nov 1999 09:17:04 -0500
Subject: Win32api.... can it create a Windows shortcut (.lnk) file?
Message-ID: <816010E2456BD111A48700805FBBE2EEFDF79E@ex-quebec-u1.baan.com>

Kevin,

There is a VB example:


      Option Explicit

      'NOTE: In Visual Basic 5.0, change Stkit432.dll in the following
      'statement to Vb5stkit.dll.

      Private Declare Function fCreateShellLink Lib "STKIT432.DLL" (ByVal _
       lpstrFolderName As String, ByVal lpstrLinkName As String, ByVal _
       lpstrLinkPath As String, ByVal lpstrLinkArgs As String) As Long

      Sub Command1_Click()

        Dim lReturn As Long

        'Add to Desktop
        lReturn = fCreateShellLink("..\..\Desktop", _
        "Shortcut to Calculator", "c:\Winnt\system32\calc.exe", "")

        'Add to Program Menu Group
        lReturn = fCreateShellLink("", "Shortcut to Calculator", _
        "c:\Winnt\system32\calc.exe", "")

        'Add to Startup Group

        'Note that on Windows NT, the shortcut will not actually appear
        'in the Startup group until your next reboot.
        lReturn = fCreateShellLink("\Startup", "Shortcut to Calculator", _
        "c:\Winnt\system32\calc.exe", "")

      End Sub

If you have NT 4, you should have Vb5stkit.dll already (I do, and I don't
have VB).
 
I think a better way with Python would be using the IShellLink COM
interface, but I have never used it.

Hope this helps,
______________________________________________________
   Gaetan Corneau
   Software Developer (Quality Assurance Team)
   BaaN  Supply Chain Solutions  
   E-mail: Gaetan_Corneau at baan.com        
   Compuserve: Gaetan_Corneau at compuserve.com  
   ICQ Number: 7395494             
   Tel: (418) 266-8252          
______________________________________________________
"Profanity is the one language all programmers know best"


> -----Original Message-----
> From: me [mailto:none at fake.net]
> Sent: Thursday, November 11, 1999 1:54 AM
> To: python-list at python.org
> Subject: Win32api.... can it create a Windows shortcut (.lnk) file?
> 
> 
> I'm just trying to find an easy to create shortcuts on 
> Windows platforms
> through Python... it doesn't seem to be a simple matter 
> though.  I've gone
> through all the Win32 docs, and found nothing, and have 
> started writing
> (with limited success) a C function to do it.
> 
> Is there an easy way?  q:]
> 
> Thanks,
> Kevin Cazabon
> kcazabon at home.com
> 
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 



From gmcm at hypernet.com  Mon Nov  8 22:15:01 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Mon, 8 Nov 1999 22:15:01 -0500
Subject: Can't use len() on returned PyList object?
In-Reply-To: <3.0.6.32.19991108211217.007cb210@mail.in-gen.net>
Message-ID: <1270005198-16611844@hypernet.com>

sessile writes:

> Why can't I use 'len()' to find the length of a list
> returned from my new extension module?  The error is
> shown as:
> 
>     TypeError: call of non-function (type int)
> 
> The list was created by 'PyList_New()' and populated
> using 'PyList_SetItem()'.  I get the same error message
> regardless of the data type held in the list.  I can
> access each element of the list just fine, but I don't
> always know how many there are.  Is there a reasonable
> way for me to find the length of these lists in Python
> other than returning the length from the extension as a
> separate object?  

Most likely you've clobbered the name "len" by binding it to an 
integer (as the error msg so tersely implies). Probably nothing 
wrong with your extension.

- Gordon



From robin at jessikat.demon.co.uk  Tue Nov  2 11:39:22 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Tue, 2 Nov 1999 16:39:22 +0000
Subject: Comparison between Python and "Ruby"
References: <1270663667-24391282@hypernet.com> <381ED7A6.D7D7FB13@compaq.com>
 <877lk1wiij.fsf@ev.netlab.co.jp> <19991102154543.A4469@optiplex.palga.uucp>
Message-ID: <alU3eRA6OxH4Ewx2@jessikat.demon.co.uk>

In article <19991102154543.A4469 at optiplex.palga.uucp>, Gerrit Holl
<gerrit.holl at pobox.com> writes
>Yukihiro Matsumoto wrote:
>>
>> I guess he loves reducing variablity, if possible, which is totally
>> against TMTOWTDI.  I don't judge whether it is good or bad.
>> 
>
>Are there people who see TMTOWTDI [1] as an advantage?
>
>[1]
>There's more than one way to do it.
>
>regards,
>Gerrit.
well I guess if the mathematica guys are right, having every known
programming language simultaneously available allows you to just learn
mathematica 
-symbolically confused-ly yrs- 
Robin Becker



From jtc at dimensional.com  Fri Nov 26 13:41:17 1999
From: jtc at dimensional.com (Jim Cochrane)
Date: Fri, 26 Nov 1999 11:41:17 -0700
Subject: Problems Finding and installing Tkinter
Message-ID: <383ED44D.3E31C959@dimensional.com>

I've been trying, with no success, to install the Tkinter package for
developing GUIs in Python.  First I tried downloading the latest version
I could find on the net, as well as the latest version of Python, to
make sure everything was in synch.  I then ran into problems with
libraries needed for Tkinter.  After downloading and installing several,
including tkinter, I still had problems and gave up.  (All of these were
.rpm files.  The last error I remember getting was for an unresovled
symbol:  __bzero.)

Next I tried uninstalling everything I had installed, and then
installing the version of Tkinter in my SuSE distribution (and
reinstalling Python).  (I'm using SuSE 6.1.)  I made sure all the needed
packages were installed, such as tk and tcl.  Again, no luck.  When
I try to execute a simple example:

#!/usr/bin/env python

from Tkinter import *
widget = Label (None, text = 'Hello GUI word!')
widget.pack()
widget.mainloop()

I get the following error message:

Traceback (innermost last):
  File "gui1.py", line 3, in ?
    from Tkinter import *
  File "/usr/lib/python1.5/lib-tk/Tkinter.py", line 5, in ?
    import _tkinter # If this fails your Python is not configured for Tk

ImportError: /usr/X11R6/lib/libtk8.0.so: undefined symbol: panic

So it's not finding a symbol named "panic".  I have no idea what library
this should be in.

Any help, clues, or pointers would be appreciated.  Also, I think
another way to do this might be to download and install the latest
version of everything needed for Tkinter - as I tried to do before - but
obtaining everything from the same place as one consistent package.
Does anyone know if this is possible and, if so, where I should look for
this?

Thanks!

Jim Cochrane
jtc at dimensional.com




From arad at fc.hp.com  Sun Nov  7 03:34:13 1999
From: arad at fc.hp.com (Arad Rostampour)
Date: Sun, 7 Nov 1999 01:34:13 -0700
Subject: win32com and IIS?
References: <802s6a$mj5$1@web1.cup.hp.com> <Q_7V3.5012$pf4.18247@news-server.bigpond.net.au>
Message-ID: <803e2a$dul$1@web1.cup.hp.com>

Thanks for the quick reply...Here is a simpler example with debugging flags
on:

import win32com.client
ieweb = win32com.client.Dispatch("InternetExplorer.Application")
ieweb.Navigate("http://www.hp.com")
print ieweb.Document.Body.innerText

Attribute Document not in cache
Getting property Id 0xcb from OLE object
Traceback (innermost last):
  File "d:\inetpub\scripts\news\news.py", line 252, in ?
    PrintHeader(form)
  File "d:\inetpub\scripts\news\news.py", line 234, in PrintHeader
    HandleMail(form)
  File "d:\inetpub\scripts\news\news.py", line 210, in HandleMail
    print ieweb.Document.Body.innertext
  File "f:\python\win32com\client\dynamic.py", line 319, in __getattr__
    raise pythoncom.com_error, details
pywintypes.com_error: (-2147023169, 'The remote procedure call failed and
did not execute.', None, None)

Again, it works from the python interactive interpreter but does not work in
a cgi script under IIS.  It doesn't like getting ieweb.Busy either...Also,
are there any issues with running the COM interface not specifically as
myself or a specific user, but rather IUSR_machinename?  Is there any
special set-up besides the Script Map key that ties a .py file to python?

Thanks,
Arad
--
Arad Rostampour- arad at fc.hp.com
NT Platform and I/O Team
Software and System Development Lab
Hewlett-Packard
Mark Hammond wrote in message ...
>I have never seen this before.  However, these details really arent enough
>to help me diagnose the problem.  It is possible to break this down
further?
>
>Mail me if you want help doing this.  A good start would be to turn on the
>debugging flags at the top of win32com\client\dynamic.py...
>
>Mark.
>
>Arad Rostampour wrote in message <802s6a$mj5$1 at web1.cup.hp.com>...
>>Hi,
>>
>>I've been using win32com to access IE and get webpages using:
>>
>>import win32com.client
>>ie = win32com.client.Dispatch("InternetExplorer.Application")
>>ie.Navigate("http://www.hp.com")
>>...
>>
>>This works fine from the command line.  But, when I try to do this from a
>>CGI script under IIS, I get the following error:
>>
>>File "f:\python\win32com\client\dynamic.py", line 319, in __getattr__
raise
>>pythoncom.com_error, details pywintypes.com_error: (-2147467259,
>>'Unspecified error', (0, None, None, None, 0, -2147467259), None)
Exception
>>exceptions.AttributeError: in ignored
>>
>>Any ideas on how to work around this?
>
>
>





From tino at wildenhain.de  Mon Nov  1 15:50:22 1999
From: tino at wildenhain.de (Tino Wildenhain)
Date: Mon, 01 Nov 1999 21:50:22 +0100
Subject: __str__ vs. __repr__
References: <19991101121709.A134039715@vislab.epa.gov>
Message-ID: <381DFD0E.26964066@wildenhain.de>

Hallo Randell,

Randall Hopper wrote:
> 
> 1) Is there a convention for what __str__ and __repr__ should return for
>    classes?
> 2) Or, whatever they return, should they return the same value?
> 3) If so, why have both in the language?
> 
>      Searching the archives yields an example with odd behavior that
> suggests they should generally return the same value:

If you test, its oviously (I thin its mentioned somewhere in the docs
too) :

__str__ represents an human-readable version of the object as string for
use
with "print" or formatet string ("%s")
__repr__ should produce a form that can be cut & pasted to construct the
named object.

print "Hello"
Hello

"Hello"
'Hello'

see the difference?
 
best regards

Tino Wildenhain



From donn at oz.net  Thu Nov 25 15:55:37 1999
From: donn at oz.net (Donn Cave)
Date: 25 Nov 1999 20:55:37 GMT
Subject: threading and timeouts
References: <943504970.076.346131773@cs.usyd.edu.au>
Message-ID: <81k7o9$n4m$0@216.39.151.169>

Quoth Piers Lauder <piers at cs.su.oz.au>:
| Are timeouts possible inside threads?
|
| Eg: escaping from an attempt to use smtplib.SMTP().sendmail() inside a
| child thread if the remote host is being extremely slow.
|
| It doesn't seem possible to use the traditional signal(SIG_ALRM,...)
| approach because signals only work inside the main thread.
|
| The only option I can think of is to re-write the smtplib module to use
| `select', but does anyone know of a simpler solution?

Note that the select option gets even more complicated if it's going
to time out the connect as well as the actual SMTP I/O.  To my vague
recollection this requires a non-blocking socket, don't think I have
ever done it myself but have seen it discussed here.

Maybe you could use two threads, the first just to start the second
which does all the work.  Meanwhile the first can "join" the thread
with a timeout, and if it hits the timeout it can do something (like
close the socket) that will insure the death of the work thread.
That last part sounds very OS-dependent to me.

Hope I'm not too confused over the threading API, haven't had any
opportunity to use it yet since the threads in my programs come out
of the OS itself.  But I guess I have to solve this same SMTP timeout
problem myself, now that you mention it.

	Donn Cave, donn at oz.net



From jeremy at cnri.reston.va.us  Thu Nov 11 11:51:46 1999
From: jeremy at cnri.reston.va.us (Jeremy Hylton)
Date: Thu, 11 Nov 1999 11:51:46 -0500 (EST)
Subject: method overriding trick
Message-ID: <14378.62498.649973.150075@bitdiddle.cnri.reston.va.us>

Overriding methods of a base class is a little clunky in Python.  You
need to explicitly name the base class that implements the method, and
you call its as an unbound method, passing self as the first argument.
Example: Base.aMethod(self, arg)

I was reading about MzScheme the other day (a Scheme variant with
object and interface support), and noticed that the MzScheme approach
translates well to Python.  MzScheme class defintions allow a rename
declarations that create a name in the class being defined that is
bound to a name in a superclass.  In Python, rename is just called
assignment. 

class Base:
    def aMethod(self, arg):
        return arg * 2

class Derived1(Base):
    super_aMethod = Base.aMethod

    def aMethod(self, arg):
        return self.super_aMethod(arg) + 1

I haven't seen this strategy for overriding mentioned before, although
I did not search the list archives.  When I described it to Guido, he
said, "Does that work?"  Has anyone used this trick before?

It doesn't solve the must-name-the-base-class problem, but it helps
manage it better.  If you keep the override assignments at the top of
the class statement, then you only have one place to look in each
class when the inheritance hierarchy changes.

Jeremy




From andres at corrada.com  Sun Nov 21 01:33:23 1999
From: andres at corrada.com (Andres Corrada)
Date: Sun, 21 Nov 1999 01:33:23 -0500
Subject: Order of keyword arguments
Message-ID: <38379233.ADE65CDB@corrada.com>

Hi,
	Can one access the order in which keyword arguments were passed to a
function? I want to call a function with the keyword syntax: f(
foo='theFooString', bar='theBarString' ) and use the fact that foo was
listed first so I can do something like:

def f( **kw ):

	Process keyword arguments remembering the order in which the were used
to call the function.
	Print first keyword
	Print second keyword
	.
	.
	.

------------------------------------------------------
Andres Corrada-Emmanuel   Email: andres at corrada.com
Owner                     http://www.corrada.com/mamey
Mamey                     Phone: (413) 587-9595
------------------------------------------------------



From psoares at consiste.pt  Wed Nov 10 12:01:15 1999
From: psoares at consiste.pt (Paulo Soares)
Date: Wed, 10 Nov 1999 17:01:15 -0000
Subject: Syntax colouring in WinEdit?
Message-ID: <51998F46F8B7D011BF7C0060B03502AD2F353E@orion.consiste.pt>

I ended up doing my own colour file. Put the following text in a file
named winedit\system\syntax\py.clr and then associate the extension py
with it and the character '#' as a comment. It doesn't work with strings
of the type '''.

***Start of py.clr***
[COLORS]
;Colors for Keywords (=1, Quoted Strings, Comments
;Default test and backgrond defined in syntax coloring dialog
;Additional colors defined here


[KEYWORDS]
and=1
assert=1
break=1
class=1
continue=1
def=1
del=1
elif=1
else=1
except=1
exec=1
finally=1
for=1
from=1
global=1
if=1
import=1
in=1
is=1
lambda=1
not=1
or=1
pass=1
print=1
raise=1
return=1
try=1
while=1


Best Regards,
Paulo Soares
> -----Original Message-----
> From:	Paulo Soares 
> Sent:	Wednesday, November 10, 1999 11:56
> To:	python-list at python.org
> Subject:	Syntax colouring in WinEdit?
> 
> Does anyone has something to do python syntax colouring in WinEdit?
> 
> Best Regards,
> Paulo Soares
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list



From donn at oz.net  Sat Nov  6 15:14:11 1999
From: donn at oz.net (Donn Cave)
Date: 6 Nov 1999 20:14:11 GMT
Subject: __str__ vs. __repr__
References: <3823E65E.D9E49CC6@python.net> <000001bf275c$02370d20$1d2d153f@tim> <7vvni7$19q0$1@nntp6.u.washington.edu>
Message-ID: <80226j$nvq$0@216.39.151.169>

Quoth "Thomas A. Bryan" <tbryan at python.net>:
...
| The problem is that this thread is the first time many of us have 
| ever thought about the problem.  I've never noticed those two lines
| in the library reference, and I haven't seen the str/repr distinction 
| clearly made anywhere else.  Furthermore, since the interactive i
| nterpreter *does* use repr(), many of us make __repr__ return the 
| human readable form so that we can easily test our classes.
|
| So, what I'm hearing is:
| 1. __str__ *should* be a human readable string
| 2. __repr__ *should* be a representation that can be eval'd back to the 
|    original object
| 3. This convention is not widely known, but
| 3b. the interactive interpreter discourages the convention by using 
|    repr to show the result of an expression

One of the problems here is that there are not two types of string
representations, but at least three.

1. Human readable representation of object.  Interpreter should display
   this in interactive use, for example.
2. Machine readable representation of object, for eval.
3. Machine readable object qua string - intentionally, not the
   object's faithful representation, but what it would be if it
   had been a string.  What the author of rfc822.Message.__str__
   appears to have had in mind, for example.

I don't see how any two of these three can be made compatible enough
to comfortably use the same implementation.  When usage parted with
documented intentions, it probably wasn't just a lot of people failing
to read the documentation.  There's a fundamental need for a machine
__str__ that's different from a machine __repr__, and the machine
__str__ is worse for general display purpose if it's intentionally
inaccurate.

Now I'm arguing that one reason many of us have never considered
this issue is that the current implementation is the best resolution
of these conflicts.  It's not completely OK, there is dissatisfaction
because people wish for the eval-able repr, but if we change course
in favor of the documented behavior, it's going to pinch somewhere else.

We already have a glimpse of this in the special cases that would
apparently be necessary.  A string's str is of course itself - string
qua string - and its repr adds quotes, and that works out fine the way
we do things now.  Change the context in which these are used - like
make str(container) use str() for contents - and you realize that the
repr was really a lot more friendly for some objects, like strings.
I'm saying this is not just a quirk of strings, it's because a string
is computable data and that's what we do use str() for.

Wouldn't it be great if that eval-able string could be shifted to
another function?  Then if you exercised that function, you could
be more or less sure to really get something the author intended to
eval with a sensible result.  Or you get an AttributeError.  Not a
string that may or may not eval, when you get around to trying it.

	Donn Cave, donn at u.washington.edu



From gmcm at hypernet.com  Fri Nov  5 12:57:07 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 5 Nov 1999 12:57:07 -0500
Subject: Problems Building Python under Cygwin
In-Reply-To: <7vtath$n1j$1@nnrp1.deja.com>
Message-ID: <1270297874-731799@hypernet.com>

salvadorej at my-deja.com wrote:

> I'm having some problems building Python under Cygwin on Windows
> NT 4.

> 2) the second problem I haven't solved. What I really wanted
> Python for was to build Zope from scratch. When I went to build
> Zope, it informed me that I needed not the "vanilla" version of
> Python, but Python compiled with thread support. OK. I go back,
> make clean, and try configure --with-threads. And partway through
> the make, I get this:
> 
>  cd Modules ; make OPT="-g -O2" VERSION="1.5" \
>                 prefix="/usr/local" exec_prefix="/usr/local" all
>  gcc  -g -O2 -I./../Include -I.. -DHAVE_CONFIG_H -c
>  ./threadmodule.c ./threadmodule.c:38: #error "Error!  The rest
>  of Python is not compiled with thread support."
>  ./threadmodule.c:39: #error "Rerun configure, adding a
>  --with-thread
> option."
>  ./threadmodule.c:40: #error "Then run `make clean' followed by
>  `make'." make[1]: *** [threadmodule.o] Error 1 make: ***
>  [Modules] Error 2

Without looking into it too deeply, you've got 2 potential 
problems here. First is that (as of earlier this year, at least) 
you could do multithreaded with mingw32 but not with 
cygwin32.

If that's been fixed, then possibly configure is not detecting 
things right. You could try editing the generated config.h and 
just #define WITH_THREAD.
 


- Gordon



From gmcm at hypernet.com  Wed Nov 24 08:05:24 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 24 Nov 1999 08:05:24 -0500
Subject: Python complaints
In-Reply-To: <867lj85g4c.fsf@g.local>
Message-ID: <1268673769-3409210@hypernet.com>

Gareth McCaughan says (among other things):

> Mark Carroll wrote:
> 
> > Basically, I'd be really interested to learn: what _don't_
> > people like about Python?
> 
>   - if/then/else works on *statements* (strictly, suites of
>     statements) rather than on *expressions*, so I can't
>     say
>       print "You scored %s point%s" % (score, if score==1 then ""
>       else "s")
>     or anything like that. (Or, as some people prefer to put it,
>     Python has no ? : operator.)

>>> def p(score):
...  print "You scored %d point%s" % (score, score==1 and ' ' 
or 's')
...
>>> p(1)
You scored 1 point
>>> p(3)
You scored 3 points
>>>


- Gordon



From tjs at longford.cs.monash.edu.au  Wed Nov 24 22:10:28 1999
From: tjs at longford.cs.monash.edu.au (Toby J Sargeant)
Date: Thu, 25 Nov 1999 14:10:28 +1100
Subject: Python complaints
In-Reply-To: <383D1AA2.35E8F0EC@compaq.com>; from Greg Ewing on Fri, Nov 26, 1999 at 12:16:50AM +1300
References: <000601bf36be$f1bf5e80$292d153f@tim> <383D1AA2.35E8F0EC@compaq.com>
Message-ID: <19991125141028.G17130@longford.cs.monash.edu.au>

On Fri, Nov 26, 1999 at 12:16:50AM +1300, Greg Ewing wrote:
> Tim Peters wrote:
> > 
> >     mutex.acquire()
> >         do stuff
> >     mutex.release()
> > 
> > is a syntax error in Python.
> 
> What's *really* needed here is a way to abstract these kinds
> of constructs. In Scheme I would arrange things so that I
> could write
> 
>    (with-mutex-acquired mutex
>       some 
>       stuff)
> 
> or in Smalltalk
> 
>    mutex do: [
>       some
>       stuff
>    ]

Just throwing this up in the air, and seeing what comes back...

class Acquire:
   def __init__(self,lock):
      self.lock=lock
      lock.acquire()

   def __del__(self):
      self.lock.release()

def function(mutex):
  _a=Acquire(mutex)
  # do stuff.

It relies on Acquire.__del__() being called as soon as _a goes out
of scope, but that doesn't seem likely to change for the moment.
(Of course doing this in JPython is a recipe for disaster... or
at least extreme inefficiency)

It also has the advantage that you can exit the function in many
ways (including by exception), and the lock will be released
regardless.

It also makes the whole function a critical section which is not
always what's wanted.

thoughts?

Toby.



From arcege at shore.net  Thu Nov  4 16:06:14 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Thu, 04 Nov 1999 21:06:14 GMT
Subject: Curses: Character in bottom right corner
References: <19991031170404.A18549@optiplex.palga.uucp>
Message-ID: <azmU3.84$AT.20506@news.shore.net>

Gerrit Holl <gerrit.holl at pobox.com> wrote:
: Hello,

: can someone explain me why the following doesn't work:

: --- *** ---
: #!/usr/bin/python

: import curses

: def main(stdscr):
:     maxy, maxx = stdscr.getmaxyx()
:     stdscr.addstr(maxy-1, maxx-1, "s", curses.A_NORMAL)
:     stdscr.refresh()

: if __name__ == '__main__':
:     import curses, traceback
:     try:
:         stdscr=curses.initscr()
:         curses.noecho()
:         curses.cbreak()

:         stdscr.keypad(1)
:         main(stdscr)                    # Enter the main loop

:         stdscr.keypad(0)
:         curses.echo()
:         curses.nocbreak()
:         curses.endwin()                 # Terminate curses
:     except:
:         stdscr.keypad(0)
:         curses.echo() ; curses.nocbreak()
:         curses.endwin()
:         traceback.print_exc()           # Print the exception
: --- *** ---

: I'm getting this error:

: 17:05:23:192/693:gerrit at gerrit:~/oefen/python/ncurses$ python test.py
: Traceback (innermost last):
:   File "test.py", line 18, in ?
:     main(stdscr)                    # Enter the main loop
:   File "test.py", line 7, in main
:     stdscr.addstr(maxy-1, maxx-1, "s", curses.A_NORMAL)
: error: [mv]waddstr() returned ERR

: It seems that he thinks it falls of the screen, but if I make either x or
: y smaller (not both!) it does work, so if I change line 7 to something like
: follows, it works.

:     stdscr.addstr(maxy-2, maxx-1, "s", curses.A_NORMAL)
:     stdscr.addstr(maxy-1, maxx-2, "s", curses.A_NORMAL)

: What am I doing wrong?

You aren't really doing anything wrong.  It's just that curses doesn't
let you write to the last character on the screen for historical (and
in some cases, practical) purposes.  On many terminals, writing to the
last character (right-most, bottom-most) causes the terminal itself to
scroll, corrupting what curses has written and what it thinks is being
displayed.  So even on terminals that allow positioning a character
there, curses will not allow it.

Try setting stdscr.scrollok(0) before doing anything else.

Wishing-he-knew-where-his-curses-book-was-ly yours,
  -Arcege




From fdrake at acm.org  Fri Nov  5 17:19:42 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Fri, 5 Nov 1999 17:19:42 -0500 (EST)
Subject: IPC8: SouthWest Airlines Internet specials
In-Reply-To: <14371.18695.714122.631868@goon.cnri.reston.va.us>
References: <Pine.WNT.4.04.9911051036270.260-100000@rigoletto.ski.org>
	<14371.14313.700685.807764@goon.cnri.reston.va.us>
	<14371.16525.555599.961527@dolphin.mojam.com>
	<14371.18695.714122.631868@goon.cnri.reston.va.us>
Message-ID: <14371.22526.178578.870222@weyr.cnri.reston.va.us>

Jeremy Hylton writes:
 > $50-$60.  She also mentioned a company called "1-800-BLUE-VAN" that
 > might offer a shuttle bus to the hotel; she was vague on details, but
 > it sounds like we have at least a phone number.

  I think I used them to catch a ride to BWI from Reston for IPC7; the 
price was $50 from Reston, so it should be cheaper to the conference
hotel from BWI.  The van was blue, at any rate.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From michael.stroeder at inka.de  Fri Nov 12 07:43:55 1999
From: michael.stroeder at inka.de (Michael =?iso-8859-1?Q?Str=F6der?=)
Date: Fri, 12 Nov 1999 13:43:55 +0100
Subject: time/strptime/mktime/localtime
References: <Pine.LNX.4.20.9911121052390.21950-100000@fep132.fep.ru>
Message-ID: <382C0B8B.DE2DC3A9@inka.de>

Oleg Broytmann wrote:
> 
> print time.strptime("1999-09-21 21:44:33", "%Y-%m-%d %T")
> 
> Linux:   (1999, 9, 21, 21, 44, 33, 6, 1, 0)
> Solaris: (1999, 9, 21, 21, 44, 33, 6, 1, 0)

A couple of weeks ago I found some oddities with strptime on my
glibc-based Linux system. Someone told me that strptime in glibc is
broken.

Ciao, Michael.



From nascheme at enme.ucalgary.ca  Tue Nov 30 00:13:40 1999
From: nascheme at enme.ucalgary.ca (Neil Schemenauer)
Date: Tue, 30 Nov 1999 05:13:40 GMT
Subject: '==' vs. 'is' behavior
References: <382C2C6A.64A8F2C2@theriver.com> <14380.12036.197083.409565@weyr.cnri.reston.va.us> <8E7F86267duncanrcpcouk@news.rmplc.co.uk> <oqemd9dk2g.fsf@titan.progiciels-bpi.ca>
Message-ID: <slrn846n92.kpg.nascheme@adler.dynodns.net>

Fran?ois Pinard <pinard at iro.umontreal.ca> wrote:
>
>By the way, is there common wisdom (or rather, maybe, usage-standards-to-be)
>about using `is' instead of `=='?

`is' will always return 0 if == returns 0 (ie. it is a finer
distinction).  `is' is slightly faster in some cases.  The only
places I use `is' is when testing if instances of objects are the
same and when testing for None.  To me, `is' means that something
is the same object.

Using `is' interchangably with == is asking for trouble.  It
works most of the time (just enough to screw you) due to
implementation details (interning of strings, caching of small
integers, etc.). 

    >>> a = 'a'
    >>> b = 'a'
    >>> a is b
    1
    >>> b = raw_input()
    b
    >>> b
    'b'
    >>> a is b
    0

Your lucky you don't do Lisp.  In scheme (r5rs) there is:

    eqv?
    eq?
    equal?

and in Common Lisp:

    eq
    eql
    equal
    equalp
    =
    char=

Python is absolute simplicity is comparison.


    Neil

-- 
python -c "f=lambda n:n and f(n/128)+chr(n%128) or '';print f(0x13b2\
e9d8829e3d1976e5dd87ae5e481e6ec3cf1e8cbb72c0eb8f0eccf879795d8f0bel)"



From tbryan at python.net  Mon Nov  8 22:02:07 1999
From: tbryan at python.net (Thomas A. Bryan)
Date: Mon, 08 Nov 1999 22:02:07 -0500
Subject: build failure for SunOS 5.5
References: <38269571.14471A7D@lag.ensieg.inpg.fr>
Message-ID: <38278EAF.1C6055E2@python.net>

Matija Exel wrote:
> 
> hello,
> 
> is there a way to build Python for SunOS 5.5? Here are my attempts
> on a SPARCstation20 platform:

Yes.  I built it at least once with no problems.  The box had a lot 
of other GNU tools available to it, but you definitely shouldn't be 
getting errors for symbols such as chown on SunOS 5.5.  I think that 
I built it with gcc, but I thought that Sun's cc was supposed to work.

Those sound like pretty severe messages in the configure.log.  Did 
./configure run to completion without error?  Can you tell which 
portion of the configure script generates the error messages?

You could always try a make clobber and start over.

---Tom



From gmcm at hypernet.com  Tue Nov 16 15:12:56 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 16 Nov 1999 15:12:56 -0500
Subject: README: Python-URL! - the weekly guide to Python resources (Nov 16)
Message-ID: <1269339321-9862524@hypernet.com>

Python-URL! - the (approximately) weekly digest of
 news and snippets from the Python community.

 For the CORBA-inclined, there was some great news out 
 of AT&T Labs, Cambridge:
  In http://www.deja.com/getdoc.xp?AN=546873219
  Duncan Grisby announced omniORB, a free C++ ORB 
  conforming to the OMG's (draft) Python mappings.
  http://www.uk.research.att.com/omniORB/omniORBpy/

 Robin Dunn announces release 2.1.11 of wxPython, 
 a wrapper for the full-featured and snappy cross-platform
 GUI wxWindows.
  http://www.deja.com/getdoc.xp?AN=549145965
  http://alldunn.com/wxPython/ 

 In a series of mysterious and understated messages
 (like http://www.deja.com/getdoc.xp?AN=548355817),
 an awesome new compendium of Python resource links is
 divulged: http://www.vex.net/~x/parnassus/. The highly
 dynamic author and maintainer can be referred to
 as "Tim Middleton", for those who are old-fashioned
 about that sort of thing.

 Jeffrey Kunce shares his configuration for doing
 a shared NT network installation of python:
  http://www.deja.com/getdoc.xp?AN=547109447
  http://starship.python.net/crew/jjkunce/ 

 There were a couple messages on Python in the big time:
  Paul Everitt in  
  http://www.deja.com/getdoc.xp?AN=546857881
  points us to an article about the merger of eGroups and  
 Onelist: 
http://www.thestandard.com/articles/display/0,1449,7543,00.ht
ml

  While Fredrik in
   http://www.deja.com/getdoc.xp?AN=546861601 
   points to an article on Searchbutton.com 

http://cnbc.com/commentary/commentary_full_story_stocks.a
sp?StoryID=6926

 Speaking of big-time, Skip Montanaro talks about the 
 configuration of one high volume Python-powered website: his
 own Musi-Cal site:
  http://www.deja.com/getdoc.xp?AN=546536571

 Finally, from the "what-the?" dept., the eff-bot comments on 
 Latin-1 and other strange characters in Python: 
  http://www.deja.com/getdoc.xp?AN=546828364

=============================================

Everything you want is probably one or two clicks away in 
these pages:

      Python.org's Python Language Website is the center of 
Pythonia
          http://www.python.org

      Much of Python's real work takes place on Special-
Interest Group
          mailing lists http://www.python.org/sigs/

      Python Consortium emerges as an independent nexus of 
activity
          http://www.python.org/consortium

      Python To-Do List anticipates some of Python's future 
direction
          http://www.python.org/cgi-bin/todo.py

      Python Journal is at work on its second issue
          http://www.pythonjournal.com

      Links2Go is a new semi-automated link collection; it's 
impressive
      what AI can generate
          http://www.links2go.com/search?search=python

    Archive probing trick of the trade:

http://www.dejanews.com/dnquery.xp?QRY=&DBS=2&ST=PS
&defaultOp=AND&LNG=AL

L&format=threaded&showsort=date&maxhits=100&groups=co
mp.lang.python


Previous - (U)se the (R)esource, (L)uke! - messages are listed 
here:
  http://purl.org/thecliff/python/url.html
or

http://www.dejanews.com/dnquery.xp?QRY=~g%20comp.lang.
python%20Python-URL%21


Suggestions/corrections for next week's posting are always 
welcome.
  http://www.egroups.com/list/python-url-leads/

To receive a new issue of this posting in e-mail each Monday 
morning,
ask <claird at neosoft.com> to subscribe.  Mention "Python-
URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to 
participate in and
sponsor the "Python-URL!" project.
-- 

Cameron Laird           
http://starbase.neosoft.com/~claird/home.html
claird at NeoSoft.com      +1 281 996 8546 FAX





From fdrake at acm.org  Thu Nov  4 17:03:42 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Thu, 4 Nov 1999 17:03:42 -0500 (EST)
Subject: C-Python tutorial?
In-Reply-To: <danU3.90$AT.20506@news.shore.net>
References: <19991101203120.A29797@optiplex.palga.uucp>
	<danU3.90$AT.20506@news.shore.net>
Message-ID: <14370.702.155351.996170@weyr.cnri.reston.va.us>

Michael P. Reilly writes:
 > The first two are on the Python website and the last is at:
 >   http://starship.python.net/~arcege/extwriting/pyext.html
 > 
 > When I get some more time here at my new job, I'll flesh this doc out,
 > get it formatted and sent off to Fred.

  I look forward to getting this when you get to it!  You can even
send it as plain text if you don't want to do the LaTeX markup
yourself.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From ivanlan at callware.com  Thu Nov 11 16:51:15 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Thu, 11 Nov 1999 14:51:15 -0700
Subject: Beginning programming with Python
References: <sng0ychgwt.fsf@motorola.com>
Message-ID: <382B3A53.5816B2@callware.com>

Hi All--

Aaron Ginn wrote:
> 
> Hello all,
> 
> I'm starting a non-profit organization dedicated to promoting computer
> literacy in low-income youth (ages 12-18). 

An admirable goal indeed.  I will be interested in hearing how well this
works.  Are you affiliating yourself with any municipal, civic or
religious organizations?

> The curriculum will
> include an introduction to programming using either Python or Perl.
> 
> First, I know Perl very well, but I know nothing about Python.  I've
> heard from many sources that Python is the best language to teach
> someone with no programming experience because it emphasizes
> fundamentals and the correct (i.e. best) way to program.  Could anyone
> comment on how the two languages compare in this aspect?  Also, I will
> need to learn Python as well if I choose to go that route.  I am very
> literate with Perl and Tcl/TK.  How quickly can I expect to pick up
> Python?
> 

*Extremely* quickly.  In fact, if you'd started yesterday you wouldn't
need to ask this question.

> Secondly, if I choose to use Python for this, I would like to supply
> each student with a reference book on Python.  I've narrowed it down
> to a choice between the two O'Reilly books, "Programming Python" and
> "Learning Python".  PP apears to have a lot more meat to it, but LP
> appears to be more suited to the beginner.  Assuming the audience
> includes kids that have never programmed before, should I go with LP,
> or will I miss out on anything by not getting the more detailed book?
> 

Of the two books mentioned, yes, _Learning Python_ would be the better
choice.  However, if you can wait, _Teach Yourself Python in 24 Hours_
is aimed *specifically* at people who have never programmed before.  It
covers:

	1)  Introduction to programming (*rock-bottom* fundamentals--what's a
variable? What's a for loop?)
	2)  Introduction to Object-oriented programming
	3)  Introduction to Tkinter

The three parts have 8 chapters each.

TYPython will be out early next year.  When is your course starting? 
Where is Chandler?-)

PP is very thorough, but needs a new edition.
LP is aimed at people who do have some idea of how to program, and is a
gentle way to further your programming skills.
Both books are funny and interesting, too.

TYPython is for total newbies; I can't speak to either funny or
interesting (modesty overwhelms me), but I try to get their attention.

None of these three books is a *reference*, however.  If you really want
a reference, then you're better off with Dave Beazley's _Python: 
Essential Reference_.  But you need to know Python before you can make
use of it.


Hope this helps,

<oh-goodness-i-am-so-shy-i-embarass-myself>-ly y'rs,
Ivan

<or-is-that-shameless-shameless-plugs-r-us?>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours
----------------------------------------------



From wtanksle at hawking.armored.net  Tue Nov  9 16:50:01 1999
From: wtanksle at hawking.armored.net (William Tanksley)
Date: 9 Nov 1999 21:50:01 GMT
Subject: What are Aspects? (Re: How can I make my assertions smarter?)
References: <Pine.WNT.4.05.9911032042440.109-100000@david.ski.org> <slrn824lqc.ngk.wtanksle@hawking.armored.net> <7vvq63$ehm$1@nntp4.atl.mindspring.net> <3826C32A.FB47EF6@compaq.com>
Message-ID: <slrn82h5o6.26j.wtanksle@hawking.armored.net>

On Tue, 09 Nov 1999 01:33:46 +1300, Greg Ewing wrote:
>Aahz Maruch wrote:

>> Sounds like templates to me, or Python's "mix-in" classes.  Am I missing
>> something?

>It seems like you can apply these things after the fact to
>already-written classes. To do the same with templates or
>mixins you would have to modify the existing code.

That's the secondary idea -- you can drop an 'aspect' class into an app
and just have it work.  This is good for Debugging, Logging, Testing, or
Optimization aspects -- the code should work both with and without them,
and it should be possible to add them without disturbing the originals.

The _primary_ idea, though, is that aspects can contain functionality
which classes simply can't.  In order to implement reference counting with
classes, for example, you have to modify tons of source code by hand, all
in seperate classes.  If you use aspects, though, you can start with code
which has _no_ memory management, and write a single aspect which
describes reference counting, and have most of your memory management
problems solved.  (A second aspect might help with cycles, and yet another
aspect to handle some memory management staticly, since in some cases it's
easy.)

>It's an interesting idea, but something about code which
>can reach out and affect the behaviour of widely scattered
>parts of the program, in ways the authors of those parts
>didn't anticipate, makes me uneasy. Smacks of the legendary
>come-from statement, somehow...

You mean like overriding a method in a subclass?  (Exactly.)  Or since
gotos are harmful, we shouldn't use any loops, because they're implemented
using gotos.

I agree with your hesitation -- AOP isn't proven yet, and right now
AspectJ looks like a limited semantic-based preprocessor more than a
language (still a nice thing to have!).  However, none of the arguments
you cite stand on their own; you need to back them up, or some smartmouth
will come up with parallel arguments to known-useful technologies.

;-)

I'm pretty impressed with aspects, especially since they're so easy to
understand.  But I see a good case for holding off on depending on them.

>Greg

-- 
-William "Billy" Tanksley



From cthuang at interlog.com  Tue Nov  9 09:25:35 1999
From: cthuang at interlog.com (Chin Huang)
Date: 9 Nov 1999 09:25:35 -0500
Subject: win32com and IIS?
References: <802s6a$mj5$1@web1.cup.hp.com> <Q_7V3.5012$pf4.18247@news-server.bigpond.net.au> <803e2a$dul$1@web1.cup.hp.com>
Message-ID: <809asv$7nq@shell1.interlog.com>

In article <803e2a$dul$1 at web1.cup.hp.com>,
Arad Rostampour <arad at fc.hp.com> wrote:
>Again, it works from the python interactive interpreter but does not work in
>a cgi script under IIS.  It doesn't like getting ieweb.Busy either...Also,
>are there any issues with running the COM interface not specifically as
>myself or a specific user, but rather IUSR_machinename?

Yes, there are issues with permissions when accessing COM objects from ISAPI
extensions.  Look at
http://support.microsoft.com/support/kb/articles/Q156/2/23.asp



From haskin at ptway.com  Tue Nov  9 16:01:05 1999
From: haskin at ptway.com (Brian Haskin)
Date: Tue, 09 Nov 1999 16:01:05 -0500
Subject: Stackless python progress?
References: <38274626.5D8C1EC2@ptway.com> <38281b64.88310413@news.oh.verio.com>
Message-ID: <38288B91.636DCE65@ptway.com>

Take a look at http://www.pns.cc/stackless/stackless.htm

This has an explanation of what it is and also some of the advantages.

Brian Haskin

Kevin Dahlhausen wrote:
> 
> I must confess I don't know this, but what is the advantage of stackless Python?
> Are you referring to the byte-code interpreter?
> 
[snip]



From mwh21 at cam.ac.uk  Wed Nov  3 14:21:33 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 03 Nov 1999 19:21:33 +0000
Subject: SWIG and function name question
References: <7vmc7e$8h$1@marcie.fissiontech.com>
Message-ID: <m3ln8f9y9u.fsf@atrus.jesus.cam.ac.uk>

kain at twilight. (Scott Barron) writes:

> Hey folks,
> 
> I downloaded and started to learn SWIG today and I ran into a little problem.
> In the C code there is a structure like this:
> 
> struct __foo { ... };
> 
> When I import my SWIG created module using 'import mymodule' I have access 
> to the struct's set/get operation via mymodule.__foo_set_member() but if I 
> import my module using 'from mymodule import *' I have access to all of my
> modules functions accept the two get/set functions that begin with __
> (I'm using dir() to check this).  This isn't too much of a problem for me
> but I'm just wondering what's going on.  Can anyone clear this up for me?

It's nothing to do with SWIG; try this for size:

put this in a file called foo.py on your PYTHONPATH:

def nowyouseeme():
    return "hi!"
def _nowyoudont():
    return "oi!"

then:

>>> from foo import *
>>> nowyouseeme()
'hi!'
>>> _nowyoudont()
Traceback (innermost last):
  File "<stdin>", line 1, in ?
NameError: _nowyoudont
>>> import foo
>>> foo.nowyouseeme()
'hi!'
>>> foo._nowyoudont()
'oi!'

Y'see, identifiers starting with `_' are considered (to some extent)
module private and are *not* imported when you use the "import *"
style.

There's a simple answer: don't use the "import *" style. Ever.

That's a bit harsh, but the only exceptions I make are for modules
that are written with that in mine, eg. Tkinter.

Also it can be handy in interactive mode, but I still tend to avoid it
there. readline rules!

HTH,
Michael

O, look: http://www.python.org/doc/current/ref/import.html has all the
gory details...



From mcfletch at vrtelecom.com  Tue Nov 30 05:56:10 1999
From: mcfletch at vrtelecom.com (Mike Fletcher)
Date: Tue, 30 Nov 1999 05:56:10 -0500
Subject: Python bugs
In-Reply-To: <001301bf3b15$b391c4d0$3acbd9c2@peridot.optichrome.com>
Message-ID: <NDBBIAPHEKGOBIHCBLIIEEBECBAA.mcfletch@vrtelecom.com>

I have build structures that are very much like that discussed.  For some
classes of application they are quite useful.  Adding an object to the
ObjectStore merely marks it as "to be stored" when the ObjectStore is
committed  and adds it to the identity cache (though a write-through flag
can override that, causing immediate committing of that item).  When the
ObjectStore is committed or destroyed it writes the current values of all
active (cached) objects to disk.  When an attempt is made to access a
variable, it is either read from the identity cache or loaded directly.

Although not as elegant as some caching schemes, it's not a truly abysmal
solution :) .  Unfortunately, that code is not mine to release, so I can't
demonstrate the value :-/ .

Enjoy,
Mike

-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Adrian Eyre
Sent: November 30, 1999 4:32 AM
To: Skip Montanaro
Cc: python-list at python.org
Subject: RE: Python bugs


>     >>> s = shelve.open("foo")
>     >>> l = []
>     >>> s["bar"] = l
>     >>> s["bar"]
>     []
>     >>> l.append("baz")
>     >>> s["bar"]
>     []
>
> I think you're asking for magic that really can't happen (at least not
> easily).
...




From phd at phd.russ.ru  Fri Nov 12 06:01:20 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Fri, 12 Nov 1999 11:01:20 +0000 (GMT)
Subject: time/strptime/mktime/localtime
In-Reply-To: <382BE48D.89CD174B@lemburg.com>
Message-ID: <Pine.LNX.4.20.9911121052390.21950-100000@fep132.fep.ru>

Hi!

   Thanks for replying.

On Fri, 12 Nov 1999, M.-A. Lemburg wrote:
> >    The following program prints "1999-09-21 21:44:33" on Linux and
> > "1999-09-21 22:44:33" on Solaris (1 hour offset).
> > 
> >    Is it a problem with my config? libs? python? Timezone is Europe/Moscow,
> > set (I think) correctly on both systems.
> > 
> > #! /usr/local/bin/python -O
> > 
> > import time
> > 
> > t = time.strptime("1999-09-21 21:44:33", "%Y-%m-%d %T")
> > m = time.mktime(t)
> > print time.strftime("%Y-%m-%d %T", time.localtime(m))
> 
> Could be that one of those sets the DST flag to 0 while
> the other sets it to 1... please provide the output of
> time.strptime() for both machines.

print time.strptime("1999-09-21 21:44:33", "%Y-%m-%d %T")

Linux:   (1999, 9, 21, 21, 44, 33, 6, 1, 0)
Solaris: (1999, 9, 21, 21, 44, 33, 6, 1, 0)

   No difference. Anyway, I think you are near to be right, as I found
there are some dates that converted back and forth correctly, without 1
hour offset.

t = time.strptime("1999-10-31 14:50:16", "%Y-%m-%d %T")
m = time.mktime(t)
print time.strftime("%Y-%m-%d %T", time.localtime(m))

   prints "1999-10-31 14:50:16", what is Ok...

   Seems the error is relative to DST. May be I have incorrect timezone
definition on Solaris...

Oleg.
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From tismer at appliedbiometrics.com  Thu Nov 11 16:05:02 1999
From: tismer at appliedbiometrics.com (Christian Tismer)
Date: Thu, 11 Nov 1999 22:05:02 +0100
Subject: ASCII delimited files
References: <942231502.1914033961@news.interbusiness.it> <80ekhb$1qrh$1@hub.org> <80et0b$2df8$1@hub.org>
Message-ID: <382B2F7E.78642C70@appliedbiometrics.com>


"D'Arcy J.M. Cain" wrote:
> 
> D'Arcy J.M. Cain <darcy at vex.net> wrote:
> > I couldn't find this on the Python or Starship sites but you can get it
> > from http://www.deja.com/getdoc.xp?AN=448769706.  One small problem with
> > it is that it doesn't handle commas within quoted strings properly.  I
> > am going to look at this today for another project and I will post it
> > here when I have something.
> 
> And here it is.  Note that Christian Tismer wrote the original (at least
> he posted it) but didn't leave his name in the script itself.  I didn't
> want to add it in since he may have had his reasons but I wanted to
> acknowledge that he wrote it.

Thanks for mentioning. FWIW, my name is of course in it
(and is still), maybe not too obvious :-)
Whem I'm posting code without any authors mentioned, then I'm always
the author. I used to forget to name myself, but never other's work.

> #! /usr/bin/env python
> # delimited.py
> # Modified from the original by D'Arcy J.M. Cain <darcy at druid.net>
> #
> # CT990226 V0.1

This says Christian Tismer wrote this crap on February 26, 1999.
Ahem, well I have to change these to be Y2K compliant! :-)

ciao - CT

-- 
Christian Tismer             :^)   <mailto:tismer at appliedbiometrics.com>
Applied Biometrics GmbH      :     Have a break! Take a ride on Python's
Kaiserin-Augusta-Allee 101   :    *Starship* http://starship.python.net
10553 Berlin                 :     PGP key -> http://wwwkeys.pgp.net
PGP Fingerprint       E182 71C7 1A9D 66E9 9D15  D3CC D4D7 93E2 1FAE F6DF
     we're tired of banana software - shipped green, ripens at home



From mikael at isy.liu.se  Wed Nov 24 05:32:21 1999
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Wed, 24 Nov 1999 11:32:21 +0100 (MET)
Subject: Python complaints
In-Reply-To: <81g8po$reo$1@mach.vub.ac.be>
Message-ID: <XFMail.991124113221.mikael@isy.liu.se>

On 24-Nov-99 Thomas Hamelryck wrote:
 >  Block delimitation by indenting. This makes python look like a porsche with
 >  a caravan attached :-). As far as I know, most python users even seem to
 >  like this for some reason.

I am one of those who appreciates this. My reason is the following:

I prefer to write readable, understandable code. To me, that means 
(among other things) indenting blocks. Hence, I already indent blocks
in every programming language, if the language allows me to. So, block 
delimiters like begin...end or {} are simply redundant to me. AFAIK, I
have not yet stumbled over a programming language that doesn't allow me 
to indent, and I would certainly avoid such a language. Block delimiting 
by indenting is actually a strong reason for me to use python. 

Someone once said: "Real programmers don't comment their code. If it was
hard to write, it should be hard to read."

I-guess-I'm-not-a-real-programmer-ly y'rs

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    24-Nov-99
Time:    11:15:28

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



From ivanlan at callware.com  Fri Nov  5 12:24:14 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Fri, 05 Nov 1999 10:24:14 -0700
Subject: Dave Beazley's Book
Message-ID: <382312BE.3913670B@callware.com>

Hi All--
I just received my copies of Dave Beazley's _Python Essential
Reference_.

Knowing Dave, I knew it would be good.  It's better than that, and New
Riders has produced an extremely nice package.

Even my wife admired it (the physical book, not the Pythonic content),
and she usually reserves her admiration for books from Yale, Harvard or
(some) Knopf.

I wish I'd had this book when I _started_ on TYPython.

<it-doesn't-hurt-that-it's-got-a-neat-colophon-either>-ly y'rs,
Ivan;-)
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From phd at phd.russ.ru  Fri Nov 12 10:28:48 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Fri, 12 Nov 1999 15:28:48 +0000 (GMT)
Subject: '==' vs. 'is' behavior
In-Reply-To: <382C2C6A.64A8F2C2@theriver.com>
Message-ID: <Pine.LNX.4.20.9911121527490.21950-100000@fep132.fep.ru>

On Fri, 12 Nov 1999, Andy Fox wrote:
> >>> a is b #are they the same object? yes.  Why?
> 1
> 
> Further experimentation shows that this seems to work for integers
> through 99, but that starting with 100 the results of the 'is' test
> return 0, as I would expect.

   I believe intergers 0 through 99 are preallocated by Python interpreter
to speed up common things like "x = y + 12".

Oleg.
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From nascheme at enme.ucalgary.ca  Tue Nov 30 03:14:36 1999
From: nascheme at enme.ucalgary.ca (Neil Schemenauer)
Date: Tue, 30 Nov 1999 08:14:36 GMT
Subject: performance of tuple-less Python?
References: <1268226961-30285235@hypernet.com> <943917954snz@vision25.demon.co.uk>
Message-ID: <slrn8471sd.m5h.nascheme@adler.dynodns.net>

Phil Hunt <philh at vision25.demon.co.uk> wrote:
>In article <1268226961-30285235 at hypernet.com>
>           gmcm at hypernet.com "Gordon McMillan" writes:
>> They're only redundant if the concept of mutability is 
>> meaningless. A reference-based language without a concept 
>> of mutability would be, um, an adventure.
>
>Why?

Some values should be immutable.  Consider if integers in Python
were mutable.  You store a reference to 1 and somewhere else the
value of 1 is changed to 2.  Is that not a problem?  Also, you
can not (efficiently) implement hash tables.

Maybe you mean that mutability is not required.  That would be an
interesting language.  How to you accomplish anything?  Sometimes
you want mutability and sometimes you don't.  That is why both
tuples and lists exist.

BTW, my example about changing the value of None is not really
the same thing.  I re-bound the global variable None to a new
value.  That is not the same as modifying the value of None.


    Neil



From akuchlin at mems-exchange.org  Sat Nov 13 14:25:04 1999
From: akuchlin at mems-exchange.org (Andrew M. Kuchling)
Date: 13 Nov 1999 14:25:04 -0500
Subject: Production Web Deployment?
References: <NBBBKMMANMGEEBHPJJECGEGIFJAA.dirkl@home.com> <3826C69F.B4E26B89@digicool.com> <14379.18122.156471.644349@ghost.bbn.com> <382D68D3.46CEFD61@digicool.com>
Message-ID: <3d4seqp52n.fsf@amarok.cnri.reston.va.us>

Paul Everitt <paul at digicool.com> writes:
> FastCGI gives us a three-year-old Apache module, plus commercial and
> supported versions for Netscape and IIS (and Apache).  Additionally,
> FastCGI has support for TCP sockets in addition to domain sockets, which
> helps deal with file permission errors when configuring it.

Indeed.  For this reason I'm not planning to go farther with mod_pcgi;
it works as a proof-of-concept, but the FastCGI approach lets me^H^H us
benefit from the work of others and not have do all the work.  

>The major reason we ditched FastCGI and came up with PCGI in the first
>place (lack of threading in the library meant a person could telnet into
>your application and lock the app) is no longer pertinent with the new
>Zope2 architecture.  This has us hoping to get FastCGI back in as a

So, does this mean that if I use the FCGIPublisher, I can have
multiple threads in Zope at the same time, without having to run
ZServer?  Good news, if so...

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
All the evils of publishing can be traced to one source -- copyright.
    -- Stefan Stykolt, quoted by Kildare Dobbs in _The Living Name_ (1964)




From dorland at lords.com  Tue Nov  2 22:39:27 1999
From: dorland at lords.com (Eric Dorland)
Date: Wed, 03 Nov 1999 03:39:27 GMT
Subject: Exceptions from callbacks
References: <381CA8E5.5D69E64@lords.com> <7virj4$b7b$1@towncrier.cc.monash.edu.au>
Message-ID: <381FAE6D.73F426FE@lords.com>

Jonathan Giddy wrote:
> 
> Eric Dorland <dorland at lords.com> writes:
> 
> >I'm writing a Python wrapper for a library. The library stores some
> >callbacks for some events, and with some glue code, get these to call
> >python functions. That's the easy part. Now since its the library these
> >callback and not Python per se, how do you pass an exception back to the
> >interpreter, because you can't return NULL from the callback. Do you
> >just set the exception, and Python will pick up on it?
> 
> Assuming these are synchronous callbacks from within the one thread (Python
> code calls C interface calls library calls C callback calls Python callback),
> you can call PyErr_Clear() before calling the library routine, then return
> if PyErr_Occurred() is true afterwards, return NULL from the C interface.
> See the Python/C API manual, section 1.3 "Exceptions".
> 
> If they happen to be asynchronous callbacks from threads started by the
> library, the simplest solution is to ignore them, maybe calling
> PyErr_PrintEx(0) to warn the user.


You're not understanding exactly. It is synchronous (ie one thread), but
since the library is calling the callback, and not the python
interpreter, I can't return NULL to the interpreter. Basically, is there
a way from inside the callback to say "Wake up python, an exception!",
without having to return NULL to the interpreter, which is impossible
from the callback.

-- 

Eric Dorland
dorland at lords.com



From m.faassen at vet.uu.nl  Mon Nov  8 12:54:14 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 8 Nov 1999 17:54:14 GMT
Subject: A Modest Proposal -- <wink>
References: <99110613580407.02970@quadra.teleo.net>
Message-ID: <8072o6$m4i$1@newshost.accu.uu.nl>

Patrick Phalen <python-list at teleo.net> wrote:
[snip]
> Please consider writing email to the HTML Working Group in support of
> this proposal.

Okay, done. A copy here:

**
Dear HTML Working Group, XML Working Group,

It is imperative and essential that you include the <wink> tag in your
proposals. If not, the Timbot will not be able to produce XHTML 
compliant output, and the whole Internet will grind to a halt as a 
consequence.

To adopt the <wink> tag some minor changes to XML need to be made. The
<wink> tag affects any _previous_ (not containing) markup. I am
proposing a system called XWTSNTSDL (XML Wink Tag Schedule Namespace
Transformation Semantic Disruption Language) to deal with this. XWTSNTSDL
depends on the two other proposals I sent to you previously: XGNBAML and
XACL (XML Grand Name But Actually MeaningLess and XML Acronym Complexification
Language).

As you can see, XWTSNTSDL is fully consistent with your current proposals. 
It is therefore expected the major internet browsers will soon be able to
support these new standards.

Regards,

Martijn Faassen, PSU (Python Secret Underground)

-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?



From gmcm at hypernet.com  Wed Nov  3 12:11:01 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Wed, 3 Nov 1999 12:11:01 -0500
Subject: C++ embeding API?
In-Reply-To: <816010E2456BD111A48700805FBBE2EEFDF74D@ex-quebec-u1.baan.com>
Message-ID: <1270473441-35788679@hypernet.com>

Gaetan Corneau wrote:
> 
> Is anyone working on a high-level C++ embedding API?
> I would be interested in that, and could help any ongoing effort
> in that direction. I have no experience writing C++ apps with
> embedded Python (I'm just beginning to read the doc), but I am an
> experienced C++ programmer.

There are at least 2 C++ libraries for Python. cxx (from the 
LLNL download) is a very sophisticated one, my SCXX is a 
vastly simpler, less complete and lighterweight one (on my 
starship pages). There may be more - check out the C++ SIG.

In both of the above, the emphasis is on making commonly 
used PyObjects look like C++ objects.

As for "embedding" - there is no embedding API. Embedding 
varies between:
  Py_Initialize();
  PyRun_SimpleString('...');
  Py_Finalize();
to highly complex interactions where you're importing 
modules, creating instances, calling methods etc. cxx covers 
a great deal more of that, but SCXX doesn't use templates and 
most of it ends up inline.

SCXX-is-probably-also-buggier-ly y'rs

- Gordon



From robin at alldunn.com  Wed Nov  3 01:25:41 1999
From: robin at alldunn.com (Robin Dunn)
Date: Tue, 2 Nov 1999 22:25:41 -0800
Subject: Need Berkley DB!!
References: <381D5E84.98931682@bibsyst.no>
Message-ID: <hAQT3.488$bu3.39846291@news.randori.com>

> After I spent most of my weekend trying to get Berkley DB to work with
> Python on my Linux ( Suse 6.1) machine, I`ve now given up completly.
> I`ve tried to install DB.1.85. and DB.2.7.7, neither worked, either
> crashing under installation ( which of course, is not Python
> related ) or when I tried to test the installation with the examples in
> the documentation ( the wrapper for the 2.7.5-module listed at
> www.python.org, under modules, databases ).

Have you tried my DB module at http://starship.python.net/crew/robind/ ?
Assuming you can get DB 2.7.x built and installed, this module should do
what you need.


--
Robin Dunn
Software Craftsman
robin at AllDunn.com
http://AllDunn.com/robin/
http://AllDunn.com/wxPython/  Check it out!







From nemeth at iqsoft.hu  Tue Nov 16 12:34:57 1999
From: nemeth at iqsoft.hu (Nemeth Miklos)
Date: Tue, 16 Nov 1999 18:34:57 +0100
Subject: Beginning programming with Python
References: <sng0ychgwt.fsf@motorola.com> <382B3A53.5816B2@callware.com> <382B3D79.74CD3E0F@email.sps.mot.com> <80p9ds$ppo$1@newshost.accu.uu.nl> <asteroidal-80psfo/INN-2.2.1/bat@broadway.news.is-europe.net> <38312D55.A36EB56D@iqsoft.hu> <antigorite-80rol2/INN-2.2.1/attentive@broadway.news.is-europe.net>
Message-ID: <383195C1.826B3342@iqsoft.hu>

Janos Blazi wrote:

> Dear Nicholas,
>
> what you say ist speculative. Your theses are not scientific predictions
> about nature. It would be very, very difficult to design an experiment that
> proves your theses wrong. Therefore from a scientific point of view, they
> are neither right nor wrong. Though they contain predictions about nature,
> these predictions should be much more carefully and precisely put to be
> really scientific.
>
> You are not saying that big project cannot be done with Python. You are
> saying that Python is less well suited for big projects than C. But it is
> not easy to atribbute a quantitative meaning to "less well". How do you want
> to scale that? You could take the number of man months necessary or you
> could try to quantify the quality of you product. Unless you provide us with
> these details I can neither confirm nor contradict. I can only check if
> there are inner contradictions in what you say. There are none, but that is
> not enough. Physics as seen by Aristotle was a wonderful piece of
> speculative thinking, without inner contradictions, bu today we know that we
> should not overestimate its practical meaning and we better take the
> Newtonian point of view.
>

Excuse me if my problem was not comprehendable. I am really not very strong in
presenting points.
The only thing I wanted, is to point out a potential shortcoming (if at all) of
Python -- I am not sure if the lack of compile-time type safety is really a
badly needed feature.
I am neither a physicist nor a mathematician, I am a simple business software
engineer. I like Python but I'd like to be convinced that Python is not only
for scientific programming, but also for business programming. I'd like to know
the general opinion of the "official" Python community on this (I think)
important topic. I learned from this list ( Fredrik Lundh) not long ago that
there is a http://www.chordate.com/kwParsing/index.html so this topic must be
hot for the Python community.

>
> Your second question concerns teaching.
>
> (1)
> I have not tried Python as a language for teaching yet so I have not had any
> experience. I should like to use Python first and give a judgement after
> using it.
>
> (2)
> I agree with you: If Python were good for nothing else, it would not be good
> for teaching either. But in ther first part of your message you only say
> that Python is not well suited for large projects (though, as i have tried
> to show you, it is not clear, what that statement means). Now you say that
> it cannot be used at all. But I have had the experience that Python is nice
> for small projects, say below 3000 lines of code. So i am aifraid your
> argumentation is vulnerable at this point.
>

As far as I understood, you are absolutely sure that Python is a perfect
language, I have to trust in it without any hesitation?

Miklos Nemeth
IQSOFT






From news at dorb.com  Mon Nov  8 07:25:50 1999
From: news at dorb.com (Darrell)
Date: Mon, 8 Nov 1999 07:25:50 -0500
Subject: Just a note
References: <381D5E84.98931682@bibsyst.no> <381D78C3.9BBECCB0@bibsyst.no> <7vut5l$pgn$1@nnrp1.deja.com> <3826A8E2.D2C557D7@bibsyst.no>
Message-ID: <LhzV3.9524$Xd2.519654@typhoon.nyroc.rr.com>

I believe Aaron was just throwing out ideas since you requirements are
vague. In most cases C++ solution will be faster than pure Python. Sounds
like you want a disk based dictionary. I'd be curious how much C++ will buy
you in this case. Disk I/O will eat into your gains.

If you try this in Python avoid readline(), but there is a fast replacement
floating around.

My current paradigm is to code it in Python then push the slow stuff into
C++. The Python phase can really clarify the problem.

I'd like to hear how you solve this.

--
--Darrell

Thomas Weholt <thomas at bibsyst.no> wrote in message
news:3826A8E2.D2C557D7 at bibsyst.no...
> > Also, please have a look at bplustree.py which
> > provides many of the advantages of berkeley db but in a pure python
> > implementation.
>
> And once again the re-occurring question when talking about pure python
> implementations of things like these :
>
> 1. speed ??
> 2. size ??
>
> Like I mentioned I`ll be storing huge amounts of data. HUGE !!! Not huge
> entries/pickled objects, but many small. It`s my impression that pure
> python approaches to problems like this ( I`m thinking of pickle vs.
> cPickle) is slow when compared to similar solutions programmed in C/C++.
>
> Thomas
>
>
> --
> http://www.python.org/mailman/listinfo/python-list





From hildeb at www.stahl.bau.tu-bs.de  Sat Nov 27 11:34:13 1999
From: hildeb at www.stahl.bau.tu-bs.de (Ralf Hildebrandt)
Date: 27 Nov 1999 16:34:13 GMT
Subject: fcntl/ioctl and O_NONBLOCK
References: <slrn83t5mv.bhk.hildeb@stahlw06.stahl.bau.tu-bs.de> <81ml5t$eer$0@216.39.151.169>
Message-ID: <slrn840205.sn4.hildeb@stahlw06.stahl.bau.tu-bs.de>

On 26 Nov 1999 18:57:01 GMT, Donn Cave <donn at oz.net> wrote:

>I don't know about the ioctl, but you may not need it after all if
>you're already going to use select.  Select will tell you if the
>I/O would block, if so then don't read.

Ah! So a read only block if there's nothing at all to read?

>(And you can't read from both ends, "stdin" (i.e., Popen3.tochild)
>is opened for write and attempts to read from it will probably be
>frustrating.)

:) Yup. I read the sockets HOWTO by now, and select() on a file/pipe 
is more or less the same on unix. 



From bparsia at email.unc.edu  Wed Nov  3 23:26:56 1999
From: bparsia at email.unc.edu (Bijan Parsia)
Date: Wed, 3 Nov 1999 23:26:56 -0500
Subject: Comparison between Python and "Ruby"
References: <1270663667-24391282@hypernet.com> <381ED7A6.D7D7FB13@compaq.com> <877lk1wiij.fsf@ev.netlab.co.jp> <19991102154543.A4469@optiplex.palga.uucp> <381F9EB4.63FA033F@python.net> <slrn81ve6m.l1c.neelk@brick.cswv.com> <382066F2.D12ECDA2@prescod.net>
Message-ID: <1e0pmag.pgoc9kfj8sd1N%bparsia@email.unc.edu>

Paul Prescod <paul at prescod.net> wrote:

> Neel Krishnaswami wrote:
> > 
> > Python (and Ruby, for that matter) remind me more of Smalltalk than
> > anything else -- ie, they have extensive libraries, a community with a
> > love of elegance but not of purity[*], and a greater focus on practical
> > applications than on academic research.
> 
> Isn't the Smalltalk community pretty enamored of purity? I mean "if" is
> an object that takes messages, right?
> 
>  Paul Prescod

Close. Boolean's are objects which take the messages ifTrue:, ifFalse:,
ifTrue:ifFalse:, ifFalse:ifTrue:, whileTrue:, etc.

All control structures are messages (typically taking blocks--aka
anonymous functions arguments):

        4 timesRepeat: [self beep]. "This is a message to an Integer."
        #(1 2 3) do: [:item | Transcript cr; show: item printString].
"do: is sent to the array (the sharp parens is the literal). The block
has one parameter, internally know as 'item'). The semi-colon means to
send the next message to the same object, i.e., Transcript. Thus
Transcript gets two messages, cr and show:."

Etc.

The only non-message type thing I can think of off-hand is the
assignment operator. Well, and the cascade.

Cheers,
Bijan Parsia.



From Gareth.McCaughan at pobox.com  Wed Nov 24 17:23:14 1999
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: 24 Nov 1999 22:23:14 +0000
Subject: Python complaints
References: <1268673769-3409210@hypernet.com>
Message-ID: <86puwz4jhp.fsf@g.local>

Gordon McMillan wrote:

[I said:]
>>   - if/then/else works on *statements* (strictly, suites of
>>     statements) rather than on *expressions*, so I can't
>>     say
>>       print "You scored %s point%s" % (score, if score==1 then ""
>>       else "s")
>>     or anything like that. (Or, as some people prefer to put it,
>>     Python has no ? : operator.)
> 
>>>> def p(score):
> ...  print "You scored %d point%s" % (score, score==1 and ' ' 
> or 's')
> ...
>>>> p(1)
> You scored 1 point
>>>> p(3)
> You scored 3 points
>>>> 

Yes. Unfortunately, as we all know, this doesn't work when
it's (say) 1 and 0 you're choosing between instead of "s" and ""...

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
sig under construction



From aaron.ginn at motorola.com  Thu Nov 11 16:12:02 1999
From: aaron.ginn at motorola.com (Aaron Ginn)
Date: 11 Nov 1999 14:12:02 -0700
Subject: Beginning programming with Python
Message-ID: <sng0ychgwt.fsf@motorola.com>

Hello all,

I'm starting a non-profit organization dedicated to promoting computer
literacy in low-income youth (ages 12-18).  The curriculum will
include an introduction to programming using either Python or Perl.

First, I know Perl very well, but I know nothing about Python.  I've
heard from many sources that Python is the best language to teach
someone with no programming experience because it emphasizes
fundamentals and the correct (i.e. best) way to program.  Could anyone
comment on how the two languages compare in this aspect?  Also, I will
need to learn Python as well if I choose to go that route.  I am very
literate with Perl and Tcl/TK.  How quickly can I expect to pick up
Python?

Secondly, if I choose to use Python for this, I would like to supply
each student with a reference book on Python.  I've narrowed it down
to a choice between the two O'Reilly books, "Programming Python" and
"Learning Python".  PP apears to have a lot more meat to it, but LP
appears to be more suited to the beginner.  Assuming the audience
includes kids that have never programmed before, should I go with LP,
or will I miss out on anything by not getting the more detailed book?

Thanks for your time.

Aaron Ginn

-- 
Aaron J. Ginn                     Motorola SPS
Phone: (480) 814-4463             SemiCustom Solutions
Fax:   (480) 814-4058             1300 N. Alma School Rd.
mailto:aaron.ginn at motorola.com    Chandler, AZ 85226



From skip at mojam.com  Mon Nov 29 16:29:04 1999
From: skip at mojam.com (Skip Montanaro)
Date: Mon, 29 Nov 1999 15:29:04 -0600 (CST)
Subject: Python bugs
In-Reply-To: <51C9F3C392D0D111A15600A0C99E64A9AF569C@localhost>
References: <51C9F3C392D0D111A15600A0C99E64A9AF569C@localhost>
Message-ID: <14402.61472.118271.508744@dolphin.mojam.com>

    Bjorn> Which is a good thing, but shouldn't preclude it from being used
    Bjorn> as a dictionary in the update statement (although the fix
    Bjorn> probably have to be to Dict.update rahter than os.environ).

You are correct.  It would have to make some assumptions about how to get a
dict from an instance object.  Should it access the instance's "__dict__"
attribute or (in this case) the "data" attribute or try to an "as_dict"
method?  Since no such well-defined interface currently exists, you're
probably better off in the short term simply executing

    d.update(os.environ.data)

    Bjorn> Cool (do you know if you can do db.items() also? -- found that
    Bjorn> one five minutes ago...)

I don't see it, and I doubt it's going to get added any time soon.  One
motivation for using on-disk mappings such as provided by bsddb and (g)dbm
is that they can be really huge (hundreds of thousands or millions of keys
are not out of the question).  Asking for values() or items() in these
situations is much more likely to get you into memory trouble than for
simple dicts which tend to be smaller.

Still, you can have this functionality easily enough:

    class MyShelf(shelve.Shelf):
        def values(self):
	    v = []
	    for k in self.keys():
		v.append(self[k])
	    return v

	def items(self):
	    i = []
	    for k in self.keys():
		i.append((k, self[k]))
	    return i

If your shelve object is really big, figure on getting a cup of coffee while
these methods execute, however...

Skip Montanaro | http://www.mojam.com/
skip at mojam.com | http://www.musi-cal.com/
847-971-7098   | Python: Programming the way Guido indented...



From a.eyre at optichrome.com  Mon Nov 29 08:50:22 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Mon, 29 Nov 1999 13:50:22 -0000
Subject: PyArg_ParseTuple() and oprtional remaining argument lists
In-Reply-To: <38422744.174116386@news.ricochet.net>
Message-ID: <000c01bf3a70$ae184ec0$3acbd9c2@peridot.optichrome.com>

> I also have trouble using PyArg_ParseTuple(...) with a single item,
> e.g.
> 	if (PyArg_Parse(args,"i",&index) && (index >= 0)) {
> is ok, but
> 	if (PyArg_ParseTuple(args,"i",&index) && (index >= 0)) {

Aha. That's the clue. Make sure that your PyMethodDef struct is like
this:

static struct PyMethodDef methods[] = {
    { "method1",   method1, 1 },
    { "method2",   method2, 1 },
    { NULL,        NULL,    0 } 
};

Not like this:

static struct PyMethodDef methods[] = {
    { "method1",   method1 },
    { "method2",   method2 },
    { NULL,        NULL    } 
};

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From timmuddletin at news.vex.net  Mon Nov 29 10:56:51 1999
From: timmuddletin at news.vex.net (tim muddletin)
Date: 29 Nov 1999 15:56:51 GMT
Subject: XML tool status?
References: <DIv04.8846$cz3.320849@nnrp1.uunet.ca>
Message-ID: <81u7o3$15sp$1@hub.org>

gvwilson at nevex.com (Greg Wilson) wrote:
>1) I can't find a Windows version of the software; and

Christian Tismer announced a Win32 installer for the XML-Sig stuff
a while back. I don't know why it hasn't been added to the XML-Sig
web page.

Anyhow, that post is at:
http://x44.deja.com/=dnc/getdoc.xp?AN=535449110

The links it gives are--
Executable binary installer:
ftp://pns.cc/pub/xml/PythonXML.EXE
http://www.pns.cc/anonftp/pub/xml/PythonXML.EXE

( I had been meaning to look this up for a while myself --- so thanks 
for the excuse to finally do it. (-: )

...



From robin at jessikat.demon.co.uk  Thu Nov  4 17:20:19 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Thu, 4 Nov 1999 22:20:19 +0000
Subject: Assorted Questions regarding the future of Python...
References: <3822f253.11537059@news.demon.co.uk>
 <002d01bf270e$238223a0$0501a8c0@bobcat>
Message-ID: <LkNjOHAjagI4EwbY@jessikat.demon.co.uk>

In article <002d01bf270e$238223a0$0501a8c0 at bobcat>, Mark Hammond
<mhammond at skippinet.com.au> writes
>[Andy writes:]
>> I haven't done any benchmarking myself, but when I installed VB6 I
>> didn't exactly leap off my seat at the speed increase.
>
>And more to the point, VB was (and is), usually, "fast enough".  Who
>does huge matrix manipulation etc in VB anyway?  Lots of the focus the
>industry has on speed is a red herring IMO.
>
>Just like Python itself - really quite "slow" compared to these
>natively compiled languages - but in the entire time I have used it, I
>have had to use the profiler exactly twice.  In every other case, it
>has been more than "fast enough" (and the development speed far
>outweighs the difference anyway).
>
>Whats-a-milli-second-between-friends-ly,
>
>Mark.
>
>
>
I did huge matrix multiplications etc. When it came to VBA 64kbytes was
the old limit in terms of easy passing to dlls. Has it gone up?
-- 
Robin Becker



From emuller at painewebber.com  Sat Nov 13 14:06:21 1999
From: emuller at painewebber.com (emuller at painewebber.com)
Date: Sat, 13 Nov 1999 19:06:21 GMT
Subject: .dll and .exe file version informaton
References: <816010E2456BD111A48700805FBBE2EEFDF793@ex-quebec-u1.baan.com> <80ha4a$pmg$1@nnrp1.deja.com> <C6XW3.10409$Xd2.618639@typhoon.nyroc.rr.com>
Message-ID: <80kcrb$tml$1@nnrp1.deja.com>

I am not a C programmer, but I did come up with a very stupid hack (see
below). This hack requires that the executable filever.exe be either in
the current path or the current directory. filever.exe is part of the
NT Resource Kit.

def FileVer(path,file):
        import win32pipe
        import string
        verdata = {}
        curDir = win32api.GetFullPathName('.')
        progPath = "FILEVER.EXE" + " /V \"" + win32api.GetFullPathName
(path) + "\\" + file + "\""
        myPipe = win32pipe.popen(progPath ,'r')
        verRaw = myPipe.read()
        # split the raw data into lines....
        lines = string.split(verRaw,"\012\011")
        #There was only basic data, parse that
        verdata['attributes'] = lines[0][0:5]
        verdata['Short OS'] = lines[0][6:10]
        verdata['Short FileType'] = lines[0][11:14]
        verdata['Short Language'] = lines[0][15:18]
        verdata['Short Version'] = string.strip(string.strip(lines[0]
[19:34]))
        verdata['Short Mode'] = lines[0][35:38]
        verdata['Size'] = string.strip(lines[0][39:49])
        verdata['Date'] = lines[0][50:60]
        verdata['Filename'] = string.strip(lines[0][61:])
        if len(lines) <> 1:
                del lines[0]
                for line in lines:
                        list = string.split(line,"\011")
                        if len(list) == 1:
                                continue
                        else:
                                verdata[list[0]] = string.strip(list[len
(list)-1])

        return verdata


If any one uses this, let me know, if anyone fixes a bug please let me
know as well.

In article <C6XW3.10409$Xd2.618639 at typhoon.nyroc.rr.com>,
  "Darrell" <news at dorb.com> wrote:
> Given the win32 calls it wouldn't be too hard to wrap them up in an
> extension module.
>
> --
> --Darrell
> <edwardm_painewebber at my-deja.com> wrote in message
> news:80ha4a$pmg$1 at nnrp1.deja.com...
> > It looks like they are not available yet....
> > In article <816010E2456BD111A48700805FBBE2EEFDF793 at ex-quebec-
> > u1.baan.com>,
> >   Gaetan Corneau <Gaetan_Corneau at baan.com> wrote:
> > > > go-ahead-just-*try*-to-write-a-program-to-do-that<wink>-ly
y'rs  -
> > tim
> > >
> > > If you really want to do it in a program, take a look at the
following
> > > functions in the help:
> > > GetFileVersionInfo
> > > VerQueryValue
> > > QueryField
> > >
>
>


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



From gerrit.holl at pobox.com  Sat Nov 27 04:35:04 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Sat, 27 Nov 1999 10:35:04 +0100
Subject: Binary vs non-binary
Message-ID: <19991127103504.A1536@stopcontact.palga.uucp>

Hello,

can someone explain me the difference in opening a file in binary mode
and opening it not in binary mode? And when is some file defined to be
binary, when it contains an ASCII 0? If so, I understand you need to say
it's binary in C, but if Python doesn't store it strings with an ASCII 0
on the end, why is there any difference?

regards,
Gerrit.

-- 
"The world is beating a path to our door"

  -- Bruce Perens, (Open Sources, 1999 O'Reilly and Associates)
 10:33am  up 17 min,  9 users,  load average: 1.81, 1.47, 0.79



From cfelling at iae.nl  Mon Nov  1 15:33:02 1999
From: cfelling at iae.nl (Carel Fellinger)
Date: 1 Nov 1999 21:33:02 +0100
Subject: Contribution: nice floating point number formatting
References: <381a4316$0$209@nntp1.ba.best.com> <19991030063317.A16991@better.net> <7vhrfm$glv$1@vvs.superst.iae.nl> <19991101124124.A903@Ridcully.home>
Message-ID: <7vktdu$u2$1@vvs.superst.iae.nl>

Malcolm Tredinnick <malcolmt at smart.net.au> wrote:
> On Sun, Oct 31, 1999 at 05:41:26PM +0100, Carel Fellinger wrote:
>> And for the spaces in the decimal part, are you shure that is conventional?
>> I can't remember having seen this *ever* in my life. At least not in my life
>> as a programmer.

> There is a convention in some situations/places where the comma (or, in
> Europe, the dot) is dropped entirely and a thin space is used (if you are
> typesetting -- something smaller than a normal inter-word space).

now that you mention the size of the space I'm inclined to think that you
are right. (I really have a rusty retension that needs to be poked now and
again:) It's also quit common these days to use such thin spaces to make
phone numbers memorizable.

> appear quite frequently. IMVHO, the spaces every three digits look nicer
> than throwing a bunch of punctuation marks at the number. :-)

I agree, but only if the spaces are trully thin, not that huge blob of white
space called a space in teletype (terminal) fonts.


-- 
groetjes, carel



From gmcm at hypernet.com  Tue Nov  2 16:40:07 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 2 Nov 1999 16:40:07 -0500
Subject: Installer v0.3d - problem with Builder
In-Reply-To: <381F0803.6D76E236@netscape.com>
Message-ID: <1270543697-31562797@hypernet.com>

Michel Orengo writes:

> I've tried to use the new beta of installer (simple) but I've got
> an Attribute Error at line 109 in 'archmodule.modules' With
> installer v0.2 (simple), it works fine. Any idea of what went
> wrong? Thanks for your help

Looks like you're talking about my stuff, but that's not enough 
for me to guess at what went wrong. If you cut & paste the 
traceback, and attach the .log file, I can probably figure it out, 
(directly to me, of course).


- Gordon



From toby at puckish.demon.co.uk  Mon Nov  8 13:31:56 1999
From: toby at puckish.demon.co.uk (Toby Kelsey)
Date: Mon, 8 Nov 1999 18:31:56 +0000
Subject: GC patch
Message-ID: <+YHjVAAccxJ4EwGk@puckish.demon.co.uk>

Here is a patch implementing garbage collection for python v1.5.2, it
compiles under linux using egcs-1.1.2.  It doesn't use the Boehm CG, or
full mark-and-sweep.  See the file gcmodule.c for details.

Currently it only checks cycles containing lists, tuples and dicts.
It has not been tested strenuously, and I would be interested in any
benchmarks people can generate.

So no-one feels left out, I have left plenty of scope for optimisation
and improvements :-).

Firstly a short script which shows what it does: 
=== cut ===
class foo:
   def __init__(s, w):
      s.greeting = w
   def __del__(s):
      print s.greeting

# list loop
a = foo('bye list')
b=[a]
del a
b.append(b)
print 'list id', id(b)
del b
# tuple-list loop
x = foo('bye tuple')
b = []
c = (b, x)
b.append(c)
print 'tuple, list ids', id(c), id(b)
del x; del b; del c
# dict loop
y = foo('bye dict')
t = {}
t[1] = y
del y
t[2] = t
print 'dict id', id(t)
del t
import gc
print 'pending', gc.pending()
print 'collect', gc.collect()
print 'pending', gc.pending()
=== cut ===

Enable DEBUG in gcmodule.c to see what is happening.

Comments, insults, free beer etc. welcome,
Toby

The patch follows...

begin 644 gc-patch.bz2
M0EIH.3%!62936:ZFK/4`&(G?@&@P>_______W\Z____^8""<`>\/ER\[O7>\
MU[/H5!5"JH.Z^VH^^^NMOGWCO:[WO!3USO7V*[=<W:[NG9KSO7>K.;N]7Q]>
M5QK.F?;[L?7NO?;3U[OOGE;9\-)\Y7UO;ZD;OO<>O/L[ON[[?`E-("`3(3%-
MJ:>A`FGD-%/$GM4!Y30T`'J:&0--!`$)IHFI/3)E,IY1M(]0`#0`````)!)$
M13U#U-/4!M0TT:,@`T``-`:```(4BD4>H]-!J>H]0#3:HTP/41@:#1J&F$8"
M#$PB2*80TF(3*>$T)HPF$FFU#3";00:#(!D`DU(28D>B9(U'II,ILF)E,&II
MD&F:@`TR`#0:#@, at 5$+W8%'LBJ01$T?_GNQP)V%45$435A&,A!23%!04K$!0
ME,110)0$3!"00[,#()IBJ*W?MVS+`WU9H=[UK4<-&EW5!!30Q2WU89!=`:_O
MFS<)42!_J7+APWFU-DL7C&/$&C((>DLB2))H*")**&D*:&)"E(*@AJ)9D):"
M@@ABB(""AD("&"(DE92`02*D0 at SZA\YVS0R/S8_-?5KX;ZWC2Q at WIG(F499)
M85#&M0N0),L9!,EF2!&E,SA:0A;!X,URQ].+9K`V-H2I6%"@0P9&B*0D"@-,
M:&;88C%IC&FTM9`NY&Y0MHS!K%F4P&HHI$O08#%2Y&@1$*H*CD-1L(4HHBB%
M)5%(H*#&BC"T%E5$VMHHD-5VQ1HB:8S;3`X+9YFT<%`3''F!2].&!NR"F)*0
MF`JH)I/04UKZC/<X8;H(MSJU!0X'/CH%R$:;AOY>8^'H_Y^Z_CL/N_GMI;_L
MPW71.S(F at C[(;H#I+C_7'E<99-]$1B=,OI+^ZG5Y_2<!3CT5QMKG`A9UI>O,
M\3ZSP<YTS8ZUN5UZX60*_#['H,6(Z4#'?\',H[?@9:=C3?##AN;A867OF(@H
M(H]VQJ(@MX.<R)LD)B$[D4NH2 at RIM[QUFSS<`]S7KRKNZ$/4!)$-T0BO<Z__
M&CIZK"BB**BF at I(DB1#H3^Q3_9Y at X>-HW]9!8R3->UHM3-ZBC#8TM^LBB:.*
M.479YFVVRAR###MK at J#3##AY%-CDBED3Q7^+,T,0]+)9^GV4:)JR:(%G,K><
M:N1)S!6*R,9$Z]CO]>\<9?=?7REZ[]SR5*E]=3@[I,G_5!DRZ.KY8Y<"5D;W
M>VY[23B99>3T\)`I#)`XC`41&""$*C0P?3\+]7L/[[]IGW]0^$?K[$P^*3A7
M].\/V?+3`JAJ,=;4>)]7LYM1B1.E#.6?4613*GSJ?-KD<)W=IV[N><,'=TC0
MH=!$^WXV=Q/C:=P_WBA!6O"'<4(([$#>[JMI8N4<FYOPL]*);YV&PT>(VO(9
MFDQ]FP2Y1Q"#>PNZ0AJ.?T.<3!J,K1;'YICN%\-F4#3;+4;1R>V^IW&^MK%[
M+8"_$]!PY<M,;F^&U]1O<R^$]7L)&[J8>PZ`YT)>U<U]K;JPE]7EN;1A.[?M
MI1%97P]2D at _$=;;R[M24Q,P%XDS)BY,0]@=9WS(UB[TNJQ'9_XD+47,:G`A<
MVK*\Z&=5'$P['MZR3[CGOU)))G86%MF,P(RN!Q%"`/SWO(L/);J?"!8(W-[%
M`2*E%0?*[;*"EO"+;U:>Z@<>K0-M`T]T>J];S6,QA5H68\\*"FK6E^V!.MM^
M0]ZY at 5Y3O;EF8,[WEKGI:L?$*J%W>[KF&IS3(%75R9V?X?);IDYFN'A01Q%?
MF/I0:/I4::P9 at DRWAG,VO3U>0L4![=AMRZW<ZTW+<6VU0VUVZM)^DQE)K*GD
MO&.&JM>F`B[F]9*DSNPW[/--1Q2MG'`N<0&3RP1=IW2M(&F?)@,=,%9WZRN&
MUZU+:A?:SRH^&[7G3B,"O*E^K.I%Q(F,RGK(79V\H;HX%2&4+]*:XTAN!0X[
M-JSR.>E8B/`LE=)]2V[GS^/<^NCL&E>!6Z$M\'\5$_%9`I-^)B-.3)&'W8>5
M>?2O6V,/9TW<*-%QMH\K[-H:SV[!]CW:SS.XT!CB&U:!%:-QB8<%,`+W,>\/
M:K3Z`5_SJ.HGS.M1UOMUN[<?D(K5"GX99::7#CDP$(#T;<$]'HN6[5`<_&NT
M`J,&'QSN+I(2[,NQ0(&"K#P!E916:UY*/L6-8Y1/!`?G01QJ\WB.:**-GIO$
M%<)B^H;*W>858`RJAX6791SI,-7JN-JFQF,\T^.7CW\I7$]VK."0>',V&41/
M8<*2GH[$X.#?B19B.E1*J"G$CQ;S$_`I>PL!/"=8,I'OPZ:B(OP;1\GCK93N
M'\$(0(9WB\.4<#(W=<CPAA29S;&O03=G33O*7K?D4W_BNV1I?B8YX780ZRI4
M#O`67+POCS4=WJOJU5L<A-`43IDHSS8GQ6YKW[K&N:"6P;5Q(&YR#0](:,E:
MB5XT/EEJ/>76G+Y<-0;^QE-%])AJGD_F#SIQUP-!MBPDJ[+2RR,.%XG)EZOU
MU%P6R-NXK at U<(F3GPL-H&M*WC4Q<J;+- at _N#-VGTV[^L at 1BVH<CB-B_411$%
MD8,!&()G6/V5=ZC`_`[G`_(XB_<D at 4!P/)8NS4^]U'1('9Z;EOM-JP,/D30Z
M^WQO9 at Z.VBR>LW<=(C)/4 at 55&3889VNTF1%,)33#]"PIYGP>?KKN2/&S[`V"
MO@"!^E/I9(0B$4U<QY"DPY]92^C'0C:ZJX;E_K#=`/?P5(S:H'2R(?9F6'(C
MY.K'(TT+6*(6#*BYOWDSN$<"BB,(K#<MK3+/!:'0)*!F!"O43H$J46*Z"X">
ML-V.&6\M"C2%0I)7O7><N4)?&!.,VWT&)?!?"C;*&]1WB6UJN4H3D4'7<.N,
MK;9Q//9\QZ<\7#I/JZ^YC6VTH.R??ATA43)%(&<?D?,\N[Y\//5*LU/!K-[I
M)$CL69!@_*+-?C\\>WYWZ.!U]H>J(6%=YFE`)C]-A$9#6C]["LH\<5NJ?!'0
M*=-`FF913C4]RR([C?[0X!ZH(OH-1NAP4Z`9B"R!X>HJE1K8YXD3VN]7$X;5
MZCAI3%%B/`BMKM!!-4V.4C>.ZA7T"\4'0:VOUEA"N67Q]%`U=FOJ':8-:+SL
M5JZA>Y6`O(,,V4QS60KCZ-*UW\;]?BU-=L<R^D+',Z8B!]9/,18,A&4I8E*5
MH.G$0/XU,0E#10>(DBA\ZJ"F at C<`X2P<A@='<+V$#`8JA1$1./I,>W,-_#]8
M5RA8#XSG)\VRT2(&.^S"ZXTX)7-:W.'HI?$WIFL"\O]0<7MJIV0]]C'*FZ?7
M2,-/%4NFX*P<2VIQ%:2(!"O95]E9-"E#NRC*LL[UM.MO'^_]W;$_6WQ/1Z>T
M/*>;[O>#C)F$]3ZI@*M"U3K`$.L^!)&S1#V'B/:+_367F?V/X#P>;SGZTBD3
MH*S(=%0^8((G6N$8^KVB3%+-RB1$U9`=8)'K3A"["XREH',=E^#HMZ@)!K&Z
M[P#IJQD3:8%FS$-[]%F'?5F,SOS6]65W/MA=.%O)INO;U_@`>,`P#,\[))I[
M.L=H8>D03UKUWHD5)$%,!898E&24.8&&600[M1$555Y!N!!#+D?P/SGS?$8D
MGO"P$UWINH^8."YEZB%3%QBJ`_DJHKU%_Z&8,)+`<D(V'.*P`ADVIV8R!!17
MD,K.OBU:8&!(^[Q`$L+G4K&#GC`_8A#VEKVZLTN;*1G4I2F(9TU`&PB:2T<+
M"Y5(1P)/(:LE$@S)FES<.&UGM/&0 at CZ(+RPP>_W&'%YJ+^?%O&U_UD.M009H
M"4/#:!DPE_2S=8%B)LK8HJ.84F7FWW+H55HFK.Z`6G9R"Z/G#`N,?C&R9`$Q
MV<W80:3+%/7%2W:B=@A+.IG!P5&095'S++R0ZBI80`*F^S<YY'U<UN/A0:C+
MZZ00U[`I6DMC`Y2357VNJ6>%X]^.85[R%N5;4G9W?U0+:OI8^&<H-OT<JLST
MFT_GT\?RG'SS\OJ_P`[_?\W7AA@^@(/"%7Q6U)E at F!TS$08"&$*TCN0P(0$$
M'UUB>:H'R),2H0`[3EPP?8$2?01.,]C/]>U#\0H+/YPX4?'[10YQHL?\8?/M
M^G0O^2=L/!J'^P0TF6H!L_FLH;VD4.";>4G^$#R'A"(&QEIP8V#0_ at VI?'>X
M<60D1.I-JDYX1YC,I+<XM`^1$CSI^(:A:AD<*YE.1""FL(VW:5NZ!K%_L`Z!
M9L^I*Z`'ZPHU`/4A@%"AY1;AM^X#57AW%3CZSU?']O9AF..,F7.4.E^V=4`'
M[XX at +Y((\1(7=8I;J)_.,]UM^^)?M0YOAY@=N`8?J+\IR^X!@8:Q$A'[9$UU
MAZG+MLQ,O-*(.#YA5DA at _:OJ]6][9;*V;:B-PAVNO.="Y_);;<!C0%#:?62!
M[6R6T!.""G%[`38,F%1!-X'6HK12`PCB1E533`X#-M0,&%V(&.GC#L0\ZF#+
M!AW*GY"AVH6,&<G(;;5-2GP^9-J>=,TWXE=P53T=5 at VL,"B2%0*(%'9=8/;^
MA*1<&)Z@][A\9(&P<]KX5JB:H(BJ)9*[POB*]ACP/+Z$P7 at 4AS)W#H@/>0Z4
MST#X&O0<H4<1;%'QB>S&+*8$H3)=_%?B"#8O7Y?A^SUO3WT$.XYWX>'W>/<Y
M=SM`9R+Z2BA:%*FN;L];J^G)VMIF#I at HK241IN6+,63&G&>CB@]E4B,U+KBA
M'T,2YL*#B'B.WX3OZDS3(*]\.LLCJ:P^,X?4J*CY_(/)3!F?J`L%U2\"Y^LP
M!88?:5$@I<F62G9T^VH at 8&GPQL2<J[SO%FWWPFCEBJ"INPN0SUGC`<W;3`LK
M:"%+[]5#C7W[<-,K3A3G)(0C6;9CRFV?NAG"R5`K?:Q=,'51MC@%/Q*16_?F
M\U$&YH;A'"U81,_SQJST_>R^X0N"](BG=T9ZB[,KSS:'9@@?`SI(OB#K#89Y
M>WF?6&_8CCD7G`<)&.-T_9,<S`+R3.J+)]#?'%@8I%T[S?'V!*VEU\WC0:X8
MY];<FECT_F88 at 2=G&"J/`*FA!XEZB:*8JB2(''#8%[)#?7[A33K%^*Y\2(TC
M#IB(P#F5J'U%S$>@^$[1I7@'`..;D(DY\S,-&N85,/$E7ID"IAXJ6MV!2`8_
M)Q^EUG)I?"KAQ@:3W=WU;E4-PY^*C.1#;A4ULB<8<<KWDA\!DD]!D#JV)YP-
M-0?XG3S:TL9A=D+#938-+]O`/?ZSG)@Z;I5#5,SMK1Y@`L36$:(4<4$H1\\G
MR2#GBDR7;1*RFG$B0<@#3TF`\8$;`+Q9XGNQJ7(PW"QK(1H9&'AEU3"FAZ=P
MG8=6H4R,VFG3)"M22(BPT*VTL5J+0R(C0Y**,J,$;[9G!)U4+6F)[&=)J0BR
MD at 4!!U'/D$YI4Z+Y[L^8K<4T']U<L^@S1S<MK9:.A+2?X#,.?LXNCQ7MO]76
MK+OF6$.4"CE$/?AIRMD`Y>0R(H`3$C?P7:O1YTGKY"TE$34]BME$011F=S3D
M1-4RP)"$+J>5!JCO";@8]K2+KB'MO)D[2D_/^%,(9$R9!]CQJ,P:%"UU30C!
MO(E`K*\X(78899S^6B2,JB4V%&%"2!$F#$YG$K:X at 8>R`8M@*)8#76TPCD2I
M*D*>5Z0-+2IB*@`4*U0X"34*()4(,&6J#'60I>VGPT:=":9160 at 3T4&;N2T0
MYB!)DNK>;H)@![3`(%'J,!)H+)F%&R-. at 0>?8N]!O%I-4?E,K!89;7;BX*F8
MXHXG_14C%A()`BN_&K"@5`/Q`B0Z#'<:.H;6K;77E"D0MNFRUM*:M/H6]ZN$
MV&SBFL42;.0-EW`Y8M1BAP8"&%E;PV[S;*K`389(*9`7>0Z3>P'W[@P#6)AC
M!D%H0%[S>SC:,R:R83IL0;=I(H&:&*P63%6;9`?3+ED)-4_)X85Z&RF,)EQ+
M<3/C!8P&RIF059-L4\`B3#24A`F`,B-*`#C1,A<%C<:72-X(572"T at 65J^2J
MP8ZI@#Q<0,$^=%39D`40F:..H*S-6]Z6Z<V689BC:%YUZ5V'L%Z[G;W_79=1
MYRE.&[NZE$.P)WQ.P["B at M"213/HM:6-LRDQF-VP+1C1U`8#!B=9%;I[V/SW
M%JX8GDKP&13A at 6`I6+18I#E[U=ZJ;1YXKST[#2YDADEJRL":XHY&0X)B0\@3
M%-G9I]H/W`1Q2M*,0 at D.74!\"X1A>/U>HM<M]SB2L/69:;#7)BYS=`DK4-HR
M8Y)U2'%#YPX'EHJ"(>=)V>%*N6)6$)91R<`P*(OT$4DD`@P^X7.`[(&+]`!!
MP'J83CC"E&#*(H.S#OE$Q`3*U`&$V=0#4'$,.8WDF/8'M$-VA,PN9DZ-P at 2_
M at +@IZ*N<)[ZEG"VI?2I,%"T1DUYEM9C@'`7PQ497(!9]:63`[+'(AZW/9GX7
M at KP)/#U)F8%F`R=O at I9'#,].C>YY6'P*(A2MH]_+8/?UQP>RP/GT/Q,QDC5E
M.1J[SQSG`A&:)Y=R=:@'[2+?Q7/JG:I)=K&!YR&9N2 at J@AO`YLA.Y[U:<^_)
MI.Y$!!B"R+MCMZ-;(-KS$#;0;#O:%A`),"1/0!P?<VM5"L<\U/HT+/H/<E0/
M6`$WYZAPWU0*H`]!1K#R#1CRTP50%#2E"D1HWU=CA@\I5HK1"#TA^(S=MX03
MD3D at G`06D:+^@N%TEVZ6M@`^G773U:%R">`OV,;%ZW&[+"[P,EKFD,C1)Q+$
M,!O>Q>%X4AIF94=`0-V1D.%6;@.@)-N.FXY;L2YL,1'X2>=GQ*R*",BA"[U!
M@%+22D#<D0T(PTKB@$D!D8=DZ at -0S"F>Z`^@$X%YM<4N&8:D at DP9,9G4(OAS
MVE!"6QX<-DK8NQ2M`S8=,RJM$+O`#<&L`S&,&3QX)/`P[XC!AAS=TK>B<_.$
MD!6EW&4!H at 07"(\!@/61`1UU#?P.*-:A39>#..(#0T<:7BIPI!T'6.Z%F!1Y
MB4J;&[[&:S5FB7_`1R:8<1H<G$`OJ8JK!A-P&$1F(:,*)%#3*64:8"[[3`#K
M"]]-=TNS*,A6H5%FI<WZ![(HTMOBH at A\;(S>/4 at RIH!C`8,!5%$$CG74PVVH
M at 0W>`S7,9`Z:>C5D=B`@[T3S"U.U;ES@(R@*J<`/`"LQ,*@Q]]EM;IHQZTVV
M\4!-941Q$94;AL$U9%A18PF?4G9W)[OS]LJP5'O)17R<+R8\U2GQ"&9PA#Q3
MB!\P)(&^/S7,V26+F?B#-#<0`A)%9!CQL>`^2)`-89J/.P,]?MR]*`6)RKQ+
MB=2>6*.'NXT!G$9B:E,'7R;B*5(1U@&]HK`;G`.*!6S`SPD.$$V^5?$^0]T-
M;\U!XTBV"+(CY2#*JJ>H5T;(4QA'E&YZ`%[UKE$<PBF^#\*I(D)I5<O5<H2>
M3NU!MP3FTZ*,T,B`=KB;$L)S;JV[^EW%*ZA"6Q>\=)O%M#/Z^0QF8=]#8Z4^
MQ"`T1%8R$CB#71`A"`D5L8)NW68#PQ-NN-SIQ*:9)!L0URHX<3 at 0;SNSH$F3
M-YA`.O+\KF3"X4E.K_8O7-VK8X9R<[S`>\8U,&28$(<X&!C0H-FB\3,9G06O
M2:[S(W%PAB9F=AU/*,9O$H;PN#D-UJ at 2BI0"P;ION6BI,C0HF42(Q81)Q(L-
MM'-@=-W'D='BBUJ08"&2&NR'NQZ49X6=W<:!\O$F-?E[A[PI?I,5SD"?.D\A
M)$4\Y at 8Y'$\GHU!CQI!8.@,$:8',D33EXN1)?R%MI"68+(HB(@E0H;?2"8!]
M3F(NFX?)YRL'CC,'<8Q<(L49A)%("2&"=5&7!$QU=9PZQ#`/F, at T#8$TDJTE
MJ.AR)SYS*^(5A3QT5,^Z^F%,*I="@5E7[=T'DGB'`K68%OC at +D/U=7*D+=@9
M+A\95*SY46%8<\ZLLS31EK!2C=MO<[8M/3XU9#F&0D!6`:DUO8<@&[U`^.VT
M:Q=BG-.A/DI507H0!.X<#Y#<6`MC6+CJJ-A1CJ-WV8.R=W628@%RD><)F330
M4U#@&_6XR=ZOWN9XSSUMMH9"$O.@@&I4OCMN;37)#:5JFX348>3E,BQO,=2(
MR!R8!5LU=Z,J8O8M(&1T#-8)K@?)%?OVF')BIGMO9US0-E)837:SOW&EFZK9
M5RSO"9)`P\KW:"0PW)',)D9D8%E at 6/R269J-OF1)*X(Y:;O"LFM[<V>Z,NQO
M&&SNA#F:E9AE+2>;L#Z8D]8GE9"B34&%"*(2>8E$-#5YIN-#,PLN`L4.*D2,
MX#04\IRW-,L at Q;CDV$IOKIXACH/)C4"!&1I&$`W@&XN9KD\&PJ6,S`LP,9`I
MJ]^5KJ2^H?*$-U`>4QH7D/#?V\!WS005(0 at 0UI:C,-$0MK-(T6%O0]*!5N`+
M6F<3$`K5/8P,:.]K!/.=TE6$TQ6<S2$"!2CCQ2=!/-U9?S.,#FQXK7>](M*@
MD1#B'O at 94`%$P)%PRY(3CP at QP,'(J%32#:%HKIR1."7"[94<>'"NE)U.TR=H
M]Y'S=U/(G_JP-Y\#<I,`#``-7Q;=1""'9518#04FH,8&XL<CVPK`9C,1)!N,
MP98%FHLV0&41T>9ODEGAO$Q$>L#74Y at X/`I)VTIT#7,SQ=<>*K)(,B<28:)#
M06F[&ZGN19L1RN8=>&3;"Q8'VJ9T6=1S]#:T0/<!$!)2%T,S2*DPR-C=>]@]
MA>02N,+R3=,,CRD(2(0*R89Q)24'H$MY3$)V0."6`UA>.NKGJ4`NA$,&`O4!
MDM\<LN.*!S/5T`QZ9:4/7CH"$IB*6:J"(\H:59#$_4<RH\&SALH!,%W"Y+#,
0 at 8=X-?N+N2*<*$A74U9Z@```
`
end

-- 
Toby Kelsey



From Gaetan_Corneau at baan.com  Mon Nov 29 16:05:17 1999
From: Gaetan_Corneau at baan.com (Gaetan Corneau)
Date: Mon, 29 Nov 1999 16:05:17 -0500
Subject: 50% Python content :)   (Distributed computing)
Message-ID: <816010E2456BD111A48700805FBBE2EEFDF862@ex-quebec-u1.baan.com>

Hello,

Sorry if my question is not strictly a Python question.

I would like to explore distributed programming, and I think Python would be
the right tool for that. I don't want to produce industrial-strenght code
right now, but learn the basics. I'm not really interested in the
"middleware" aspects, so I didn't subscribe to the "Python Distributed
Objects" SIG right away. I just want to answer questions like how do I
partition a monolithic program to run on multiple computers?

I went to the bookstore, and found some title (not with Python), but I
wasn't sure which one to buy. Those are expensive, and I thought that
someone here could point me in the right direction. I'm looking for
practical, hands-on stuff, if possible.

Thanks in advance,
______________________________________________________
   Gaetan Corneau
   Software Developer (Quality Assurance Team)
   BaaN  Supply Chain Solutions  
   http://www.baan.com
   E-mail: Gaetan_Corneau at baan.com                  
   Tel: (418) 266-8252          
______________________________________________________
"Profanity is the one language all programmers know best"




From king at calibre-dd.com  Wed Nov 24 12:15:35 1999
From: king at calibre-dd.com (Stephen J King)
Date: Wed, 24 Nov 1999 12:15:35 -0500
Subject: Numeric module under IRIX 6.5
Message-ID: <383C1D37.41863D9B@calibre-dd.com>

I've downloaded LLNL Numeric module and compiled it for IRIX 6.5 (using
-n32 flag), but when importing the module to Python, it dumps core |=(

Has anyone encountered/solved this problem?

Thanks

--  Stephen J King    Technology Manager   Calibre Digital Pictures  
--  king at calibre-dd.com               http://www.calibredigital.com



From gmcm at hypernet.com  Thu Nov  4 19:41:05 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Thu, 4 Nov 1999 19:41:05 -0500
Subject: NT DLL modules and C++
In-Reply-To: <7vqf95$5rg$1@sun500.nas.nasa.gov>
Message-ID: <1270360037-42609785@hypernet.com>

Patrick J. Moran writes:

> I should clarify a bit.  I have an initFOO() call for my FOO
> module, and it gets called just fine.  I can manually call
> "placement new" constructors on the static data members of my
> classes to initialize them, but it's awkward and a hack.  It's
> also not a complete solution because there may be other static
> objects that belong to libraries whose implementation I'm not
> familiar with, e.g., iostreams.  There's got to be a better way.
> 
> My understanding is that there is an entry point,
> _DllMainCRTStartup, that the linker generates that calls the
> necessary C++ constructors, and then calls DllMain.  

It's a routine in the C runtime, not in your DLL.

> So, the good
> news is that I could call _DllMinCRTStartup and the appropriate
> construction code would be called, assuming I figure out the
> arguments required by _DllMainCRTStartup. 

I don't think you can call it.

> The bad news is that
> _DllMainCRTStartup calls DllMain, the entry point provided by the
> user (or the Python library, I presume).  

DllMain can be provided by the dll author. If he/she doesn't 
provide one, a dummy one will be created.

> DllMain in turn calls
> my initFOO (again, I presume) so I would be in an infinite loop. 
> Of course I could think of hacks to prevent the infinite looping.
>  Still, there's got to be a better way.

Nope. DllMain (the default one) does nothing. If you write one, 
it's generally to do things like set up TLS (which you probably 
_don't_ want to do with Python).
 
> The solution that I'm looking for is to get Python to call the
> constructor code when it does the dynamic linking.  I think this
> means calling _DllMainCRTStartup rather than DllMain.  I've tried
> looking through the doc in the PC directory, but it's still
> unclear to me.  That's what I was looking for help with.

It's not Python's responsibility, it's the runtime library.

I haven't tried it in a couple years, but I was never successful 
in getting static initializers to work with DLLs. In fact, I had a 
lot of problems with them in normal MSVC apps, so I got out 
of the habit of using them. I just "new" them on startup.

- Gordon



From T.Middletin at news.vex.net  Mon Nov 15 13:19:42 1999
From: T.Middletin at news.vex.net (T. Middletin)
Date: 15 Nov 1999 18:19:42 GMT
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <JlFX3.3659$s87.67930@dfiatx1-snr1.gtei.net> <80oqn9$ocq$2@newshost.accu.uu.nl> <oLUX3.286$yA4.4821@dfiatx1-snr1.gtei.net> <80p9su$ppo$2@newshost.accu.uu.nl>
Message-ID: <80piru$13rl$1@hub.org>

On 15 Nov 1999 15:46:38 GMT, Martijn Faassen <m.faassen at vet.uu.nl> wrote:
>mature enough, Parnassus can have a flag somewhere 'distutils compatible'
>or something like that for each module/program in its database.

Yeesh! Got about a hundred times more hits on my web site today than i've
ever gotten. And now i'm posting ten times more usenet messages in one
day than i usually do in a week or a month or more! Grumble grumble. (-;

Er... i mean, nice idea. (-: I'm game for it. Anything to make 
distributing python stuff easier, i'm interested in definitely.
Will have to keep this in mind and see if it goes anywhere...

...



From dkaznadzey at my-deja.com  Mon Nov  8 14:44:16 1999
From: dkaznadzey at my-deja.com (dkaznadzey at my-deja.com)
Date: Mon, 08 Nov 1999 19:44:16 GMT
Subject: Popup menus in Tkinter
References: <199911081726.SAA25412@axil.hvision.nl>
Message-ID: <80796f$f9h$1@nnrp1.deja.com>

Thanks A LOT !
It worked perfectly!
What would be a good source of information on tk commands
available through Tkinter?

Denis


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



From bernhard at alpha1.csd.uwm.edu  Sat Nov 27 14:57:27 1999
From: bernhard at alpha1.csd.uwm.edu (Bernhard Reiter)
Date: 27 Nov 1999 19:57:27 GMT
Subject: XPython (X11)
References: <383E5350.1E07536A@xrce.xerox.com> <slrn83tj53.k7h.bernhard@alpha1.csd.uwm.edu> <81ms2n$126$1@essle.valdyas.org>
Message-ID: <slrn840dt7.mov.bernhard@alpha1.csd.uwm.edu>

On 26 Nov 1999 20:54:47 GMT, Boudewijn Rempt <boud at rempt.xs4all.nl> wrote:
>Bernhard Reiter <bernhard at alpha1.csd.uwm.edu> wrote:
>> On Fri, 26 Nov 1999 10:30:56 +0100, Emmanuel Pietriga <emmanuel.pietriga at xrce.xerox.com> wrote:

>> If I want to program a GUI I look in to wxPython.
>> Or pyGtk or pyFOX or pyFLTK, but XPython... :)

>Hey! Don't forget PyQt and PyKDE (and PyQt seems to be
>working under Windows, too, nowadays)...
Do you have to but the Troll Kit first? >:)

And I didn't mention pyGNOME, why should I mention to PyKDE?
To create an application tied to KDE or GNOME is a mistake in my
eyes anyway. <wink>

people-will-love-it-any-way-ly, yours
	Bernhard
-- 
Research Assistant, Geog Dept UM-Milwaukee, USA.  (www.uwm.edu/~bernhard)
Free Software Projects and Consulting 		         (intevation.net)  
Association for a Free Informational Infrastructure            (ffii.org)



From mwh21 at cam.ac.uk  Tue Nov 16 06:16:12 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 16 Nov 1999 11:16:12 +0000
Subject: exec/eval changed?
References: <942690155.686795566@news.glph1.on.wave.home.com> <m34sen35vi.fsf@atrus.jesus.cam.ac.uk> <3830AC4B.1C8C1107@kw.igs.net>
Message-ID: <m3zowe1ybn.fsf@atrus.jesus.cam.ac.uk>

JB <jimbag at kw.igs.net> writes:

> Michael Hudson wrote:
> > 
> > JB <jimbag at kw.igs.net> writes:
> > 
> > > Did the exec/eval statements change somehow in the 1.5 versions? I
> > > have a script that did an eval( "some code", dict1, locals()) but
> > > now it doesn't work. I've looked through the doc's but they are
> > > little thin as to these statements. Any help is appreciated.
> > 
> > Well, eval("some code") is a syntax error. What are you trying to to
> > do? We can't read minds here.
> > 
> > Regards,
> > Michael
> 
> My apologies. I was thinking that the scoping for the 2nd and 3rd
> parameters to the eval call had changed somehow and thus the code was
> irrelevant to the example.
> 
> What I am trying to do is simple eval a string with python formatting
> operations within and return the result. The code string is dynamic.
> Basically: 
> 
> eval( "'The value of t is %(t)s' % t", self.private_dict, locals())
> 
> where t is in self.private_dict. The questions comes because this code
> worked in python v1.5 but does not work in v 1.5.2 and I was curious if
> there had been any changes to the scoping or to the eval/exec calls
> themselves.

How does it fail? If I try something like that I get:

Traceback (innermost last):
  File "<pyshell#3>", line 1, in ?
    test(2).go()
  File "<pyshell#2>", line 6, in go
    return eval( "'The value of t is %(t)s' % t", self.private_dict, locals())
  File "<string>", line 0, in ?
TypeError: format requires a mapping

If I change it to

return eval( "'The value of t is %s' % t", self.private_dict, locals())

then it works.

You aren't trying to do 

return 'The value of t is %(t)s'%self.private_dict

are you? The use of eval seems a bit overkill here!

> Sorry for the confusion.

That's OK; but for future reference you might want to include how your
problem code fails as well!

HTH,
Michael



From cgw at fnal.gov  Thu Nov 11 10:11:30 1999
From: cgw at fnal.gov (Charles G Waldman)
Date: Thu, 11 Nov 1999 16:11:30 +0100 (MET)
Subject: print "foo", without a space
In-Reply-To: <m3u2mtoyv3.fsf@atrus.jesus.cam.ac.uk>
References: <382AD3BB.69899CBD@helsinki.xx>
	<m3u2mtoyv3.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <14378.56482.32270.846010@gargle.gargle.HOWL>

Michael Hudson writes:
 > Mikael Johansson <mikael.johansson at helsinki.xx> writes:
 > 
 > > Hello All!
 > > 
 > > Is there another way of getting print to not make a newline than
 > > using:
 > >     print "foo",  ?
 > 
 > I think there's some way invloling sys.stdout.softspace, but I'm no
 > sure what. However...

Another option is to forego the "print" and use "sys.stdout.write()"
instead.  Conversion of the argument to a string is your
responsibility, and no extra spaces/newlines/etc are added to the
output.




From tbryan at python.net  Thu Nov 11 21:35:33 1999
From: tbryan at python.net (Thomas A. Bryan)
Date: Thu, 11 Nov 1999 21:35:33 -0500
Subject: Beginning programming with Python
References: <sng0ychgwt.fsf@motorola.com>
Message-ID: <382B7CF5.68048F5F@python.net>

Aaron Ginn wrote:
> 
> Hello all,
> 
> I'm starting a non-profit organization dedicated to promoting computer
> literacy in low-income youth (ages 12-18).  The curriculum will
> include an introduction to programming using either Python or Perl.

I'd definitely vote for Python.  I've programmed in both quite a bit, 
and I can't imagine using Perl over Python for this purpose.  Some things 
that spring to mind:
1. Python's interactive interpreter will let them get their feet wet 
   immediately and will continue to help them test code as they progress.
2. Multiple IDEs exist for programming in Python.
3. Python's core syntax is much simpler than Perl's.  Learning some of 
   Python's libraries will eventually be useful, but students can do a 
   lot of learning before even dipping into the library.
4. Python's design nudges programmers toward what I would consider a better 
   style and overall architecture.  I believe that coding in Python improved 
   my programming in all other languages that I use.  I think that some of 
   its design (indentation-based blocks, any file is a module, convenient 
   object-oriented support) make would really help students get off to a 
   good start.  
5. Objects and references are simpler to explain and use in Python than in Perl.

See also the "Making the Python case to my dept." thread from a few 
days ago.  The thread there was geared toward using Python at a University, 
but many of the same arguments apply in your case, too.

---Tom



From Gareth.McCaughan at pobox.com  Wed Nov 24 05:38:27 1999
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: 24 Nov 1999 10:38:27 +0000
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk>
Message-ID: <867lj85g4c.fsf@g.local>

Mark Carroll wrote:

> Basically, I'd be really interested to learn: what _don't_ people like
> about Python?

  - if/then/else works on *statements* (strictly, suites of
    statements) rather than on *expressions*, so I can't
    say
      print "You scored %s point%s" % (score, if score==1 then "" else "s")
    or anything like that. (Or, as some people prefer to put it,
    Python has no ? : operator.)

    In fact, the distinction between statements and expressions
    is in my view a wart. (Lots of people would, of course,
    disagree.)

  - Memory management is by refcounting instead of GC. This
    has its good points (it means that one can define more
    precisely when e.g. finalisation happens), but it does
    mean that if you build structures with cycles in the
    reference graph then they won't get GCed when you want
    them to. This is a real nuisance, for instance, when
    you're exploring ideas or debugging: suppose you have
    a piece of code that builds a big data structure out of
    class instances; then every time you make a small change
    and run it again (you have to run it again, because
    existing instances belong to the old versions of the
    classes) you can be leaking a lot of memory.

  - Method definitions have to be lexically contained in
    class definitions. So suppose you write something that
    parses a language and then does various things with
    the parse tree; if you adopt an OO approach to the
    parse tree (with things like "class IfStatement(ParseNode):")
    then you can't separate out very different operations like
    foo.print(), foo.compile() and foo.evaluate(). (That is,
    you can't group all the print methods together, and group
    all the compile methods together somewhere else, etc.)
    You can, of course, split your code up by having the WhileLoop
    and UntilLoop classes in different files, but this seems
    less sensible to me. :-) (This particular gripe applies to
    most OO languages.)

  - The Python interpreter is rather slow. (I'm talking about
    CPython here. I believe JPython is usually somewhat slower.)

  - Some more iteration constructs would be nice.

      - do ... while
      - parallel iteration across more than one list
      - a less yucky way of saying "for i in range(1,20,3)"

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
sig under construction



From malcolmt at smart.net.au  Tue Nov 30 19:17:51 1999
From: malcolmt at smart.net.au (Malcolm Tredinnick)
Date: Wed, 1 Dec 1999 11:17:51 +1100
Subject: What's the canonical vi setting for Python indentation
In-Reply-To: <Pine.LNX.4.20.9911301540450.30390-100000@seanb.sttln1.wa.home.com>; from Sean Blakey on Tue, Nov 30, 1999 at 11:46:10PM +0000
References: <38445B6A.A1283D2E@cs.mu.oz.au> <Pine.LNX.4.20.9911301540450.30390-100000@seanb.sttln1.wa.home.com>
Message-ID: <19991201111751.A1277@Ridcully.home>

On Tue, Nov 30, 1999 at 11:46:10PM +0000, Sean Blakey wrote:
> I don't know about canonical, but I have the following in my .vimrc
> set expandtab	"Turn's tabs into spaces
> set hardtabs=4	"Make tabs 4 spaces wide
> set tabstop=4
> set shiftwidth=4	"For use with << and >>

I tend to avoid 'set expandtab' to save typing. If I want to back out from one
indentation level to the previous one, it takes a single BackSpace to do so.
If I am using the 'expandtab' setting, it takes four keystrokes (these things
matter to me). Of course, I am also using the 'autoindent' setting (but, then
again, doesn't everybody in vim?).

To avoid making this a complete "me too" post, I'll point out that the "retab"
command is useful at some points, too. After I've edited a program to more or
less completeness, if I'm going to email it or print it, I will do
   :se et
   :1,$ret

You can go back the other way by doing
   :se noet
   :1,$ret!   <--- Changing sequences of spaces back to tabs

This way I get the best of both worlds -- efficient typing speeds, but
something that doesn't rely on the receiver's (printer or email adressee) tab
settings to look the same as my on-screen copy.

Cheers,
Malcolm Tredinnick

-- 
42.7 percent of all statistics are made up on the spot




From jblazi at netsurf.de  Tue Nov 16 16:17:01 1999
From: jblazi at netsurf.de (Janos Blazi)
Date: Tue, 16 Nov 1999 22:17:01 +0100
Subject: Beginning programming with Python
References: <sng0ychgwt.fsf@motorola.com> <382B3A53.5816B2@callware.com> <382B3D79.74CD3E0F@email.sps.mot.com> <80p9ds$ppo$1@newshost.accu.uu.nl> <asteroidal-80psfo/INN-2.2.1/bat@broadway.news.is-europe.net> <38312D55.A36EB56D@iqsoft.hu> <3831A4EB.76FF32AF@boi.hp.com>
Message-ID: <acquit-80shkh/INN-2.2.1/a&m@broadway.news.is-europe.net>

There is one thing though in Python I do like at all!
The language is not as versatile as PERL when loops are concerned. I very
much like the construction

  REDO ... UNLESS

or something like this. It is a pity Python does not have this sort of
things.

Janos B.



Bill Anderson <banderson at boi.hp.com> schrieb in im Newsbeitrag:
3831A4EB.76FF32AF at boi.hp.com...
> Nemeth Miklos wrote:
> >
> > Janos Blazi wrote:
> >
> > > I have tried to use Perl in my teaching. It has a very powerful but
alas!
> > > not very simple syntax! To start with, my pupils used to forget the
dollar
> > > signs and Perl does not like that. So this year I shall use Python. We
shall
> > > see. Time will tell, if Python is really better. Maybe my pupils will
forget
> > > the indenting this time? There is actually nothing else you can forget
in
> > > Python.
> >
> > I am quite new in Python programming, and I was a bit :-( by
understanding that
> > Python does not have any (even optional) mechanism to enforce the usage
of
> > predeclared variables and functions. The complete lack of compile-time
> > type-safety of Pythom may be a problem in large projects, which may be
the most
> > serious technical (ie non-marketing) obstacle to Python's becoming a
widespread
> > language like Java.
> > Let us see an example:
> > # Module t.py
> > def f():
> >   print "started"
> >   x = 128
> >   x.foo()
> >
> > Let us import it in the Python interpreter:
> > >>> import t
> > >>>
> > You can see thet the interpreter compiled t.py to t.pyc (or t.pyo) but
found no
> > error.
> > All typos (typing errors of variable and function names) will be
detected ONLY
> > at runtime! Python is perfect language to write programs fast, but with
the
> > cost of hordes of testers.
>
> Disagree. I write many things in python, fast, and without hordes of
> testers.
>
> > However this may not be a big problem, because any
> > piece of software should be thoroughly tested, to make sure that there
are no
> > semantic errors in the application. In the case of Python the
applications
> > should also be tested for "syntactic" errors.
>
> As should _any_ language.
>
> > In Python (just like in Smalltalk or in Perl, or in Tcl) there are a lot
of
> > things a programmer may forget.
>
> Just like in C, java, C++ ...
>
> > I am on my way of becomming a Python programmer, but this should not
prevent me
> > from seeing the pitfalls of the language.
> > A tutor must emphasize these gotchas of any programming language.
> > Perl has a facility called 'use strict': it is also very poor compared
to C++
> > and Java strict static (ie compile time) system.
> > The only really large project implemented in Python I found is Zope.
Zope is a
> > fantastic piece of software, and may be regarded as an evidence of my
being
> > totally wrong ;-).
>
> yup :-)
>
> > > I feel that it is the right language for teaching.
> >
> > Anyway, Python is a great language for teaching, but why to learn a
language,
> > which is good only for teachning?! This is the real question you (or the
Python
> > community) should answer!
>
> Not possible. The question is predicated upon an assumption. You
> incorrectly assume that Python is _only_ good for teaching. There are
> quite a number or large companies that would disagree with you.
>
> Bill
>
>
> --
> Bill Anderson Linux/Unix Administrator, Security Analyst
> ESBU (ARC) banderson at boi.hp.com
> My opinions are just that; _my_ opinions.





From billtut at microsoft.com  Wed Nov 24 09:37:24 1999
From: billtut at microsoft.com (Bill Tutt)
Date: Wed, 24 Nov 1999 06:37:24 -0800
Subject: Python and COM: a bug?
Message-ID: <4D0A23B3F74DD111ACCD00805F31D8100DB915AA@RED-MSG-50>


> From: Gaetan Corneau [mailto:Gaetan_Corneau at baan.com]
> Here is my original post:
> 
> 
> Hello,
> 
> I have encountered a weird problem a few times in the past months. I'm
> testing COM server (written here) with Python, and it seems 
> that there is
> something wrong in the code that optimizes the calls to COM objects.
> Sometimes (a very rare occurence) the wrong function gets 
> called when I
> access a property. Suppose I have object O with properties a 
> (an int), b (a
> date), c (a string), then
> 
> print O.b
> 
> would give me 
> "Hello, world"
> 
> The error always seem to be off by one (in the same 
> direction), but I'm not
> absolutely sure about that.
> 

How is this "O" written?
Is it an ATL C++ COM object that uses IDispatchImpl? (i.e. a dual interface)
Is it a C++ COM object that implements IDispatch on its own? (not a dual
interface)
Is it a C++ COM object that uses OLEAUT's builtin IDispatch, but just
doesn't use ATL? (i.e. a dual interface)

My hunch would either be a bug in Python dispid code, or your COM object has
a bug in some custom IDispatch code.

I also don't think we try to optimize any calls. COM doesn't give one enough
information to even bother optimizing anything other than caching Name ->
DispID mappings.

> And sometimes, I get something like that:
> 
> >>> Object
> <COMObject <unknown>>
> >>> P = Object.Parent # Parent is an attribute that is the object that
> created the current object
> >>> P
> <method CDispatch.Parent of CDispatch instance at c5e780>
> >>> P = Object.get_Parent()
> >>> P
> <COMObject <unknown>>
> >>>
> 

What really helps in cases like this is to have the IDL for your
dispatch/dual interface so that we can examine whats happening much easier.
:)

The otherthing to note is that the <COMObject <unknown>> occurs because you
aren't wrapping the typelibrary with Python helper classes. Using
win32com.client.EnsureModule("<typelib guid>") allows you to support a
couple of features that are impossible to deterimine without the typelib. I
forget exactly what those are off the top of my head.


> I have workarounds, but what really annoys me is that I'm 
> still fighting a
> couple of guys here to get Python accepted: they always ask 
> "why don't you
> just use VBScript?", and it's true that I don't have those 
> problems with
> VBScript (because VBScript doesn't try to optimize calls).
> 

The answer believe it or not is, "Python is the most powerful scripting
language on the planet for Win32 programming." its multi threaded, COM
happy, ActiveX Scripting Engine happy, and really easy to talk to. VBScript
on the other hand... Don't get me started... :) 

> I'll try to find what the problem is by tracing the Python 
> code, but since
> I'm not familiar with it, I don't expect to find the cause. 
> At this point,
> all I want to know is if I'm the only one with that kind of strange
> behavior...
> 

I've seen behavior like the above "Parent" issue before. 
I haven't ever seen the first behavior you mentioned...

Bill



From skip at mojam.com  Fri Nov  5 15:39:41 1999
From: skip at mojam.com (Skip Montanaro)
Date: Fri, 5 Nov 1999 14:39:41 -0600 (CST)
Subject: IPC8: SouthWest Airlines Internet specials
In-Reply-To: <14371.14313.700685.807764@goon.cnri.reston.va.us>
References: <Pine.WNT.4.04.9911051036270.260-100000@rigoletto.ski.org>
	<14371.14313.700685.807764@goon.cnri.reston.va.us>
Message-ID: <14371.16525.555599.961527@dolphin.mojam.com>

    Jeremy> Everyone planning to fly to the conference should consider
    Jeremy> flying through BWI.  I have flown from BWI to San Jose and San
    Jeremy> Francisco a few times this year, each time for about $200.
    Jeremy> Flights through BWI are often cheaper than flights from Dulles
    Jeremy> or National.

If you're on a budget and might like to get away without renting a car,
don't forget to figure in the cost of ground transport into the total
equation.  For IPC1 I flew into BWI and found the metro didn't go there.
Not real convenient.  At least with National you've got a Metro station
nearby.  

All that said, I don't know how convenient the conference location is to
public transport.  Jeremy, Guido, and other locals (yikes! Guido's a local
now?) can probably give a summary of that sort of thing.

me-who-rarely-drives-a-car-to-work-ly y'rs

Skip Montanaro | http://www.mojam.com/
skip at mojam.com | http://www.musi-cal.com/
847-971-7098   | Python: Programming the way Guido indented...



From gmcm at hypernet.com  Fri Nov 19 09:06:19 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 19 Nov 1999 09:06:19 -0500
Subject: Python in business programming [was Re: Beginning programming with Python]
In-Reply-To: <19991119135811.B4306@optiplex.palga.uucp>
References: <1269336472-10033930@hypernet.com>
Message-ID: <1269102119-24129916@hypernet.com>

Gerrit Holl wrote:
 
[snip] 
> I agree. But some functions are too small to be used:
> in os.path:
> def os.path.isabs(s):
>     return p[:1] == '/'
> 
> I think that's a redicolous (how does one spell that?)
> function...

"Ridiculous" on *nix, "Wrong" on Windows.

- Gordon



From phd at phd.russ.ru  Sat Nov  6 08:39:21 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Sat, 6 Nov 1999 13:39:21 +0000 (GMT)
Subject: Get parameter by name
In-Reply-To: <m366zfai1u.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <Pine.LNX.4.20.9911061339050.18974-100000@fep132.fep.ru>

On 6 Nov 1999, Michael Hudson wrote:
> locals() should do the trick:
> 
> def f(x):
>     return locals()["x"]

   Thanks!

Oleg.
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From Gaetan_Corneau at baan.com  Fri Nov 12 11:53:58 1999
From: Gaetan_Corneau at baan.com (Gaetan Corneau)
Date: Fri, 12 Nov 1999 11:53:58 -0500
Subject: .dll and .exe file version informaton
Message-ID: <816010E2456BD111A48700805FBBE2EEFDF7AE@ex-quebec-u1.baan.com>

> Given the win32 calls it wouldn't be too hard to wrap them up in an
> extension module.

Yes. The only problem is that the file information is divided in two parts:
the first one is fixed, and is independant of locale, the second is
user-defined. The second part is harder to read.

If anyone is interested (is that really useful?), I could code the fixed
part and publish the C++ code. That would require only a few minutes of my
time.

______________________________________________________
   Gaetan Corneau
   Software Developer (Quality Assurance Team)
   BaaN  Supply Chain Solutions  
   E-mail: Gaetan_Corneau at baan.com        
   Compuserve: Gaetan_Corneau at compuserve.com  
   ICQ Number: 7395494             
   Tel: (418) 266-8252          
______________________________________________________
"Profanity is the one language all programmers know best"




From victor_ng at my-deja.com  Mon Nov 15 00:32:59 1999
From: victor_ng at my-deja.com (victor_ng at my-deja.com)
Date: Mon, 15 Nov 1999 05:32:59 GMT
Subject: Debian/Apache problems
Message-ID: <80o5ub$csn$1@nnrp1.deja.com>


<Directory /usr/lib/cgi-bin/python>
PYTHONPATH /usr/lib/cgi-bin/python/stdlibs
</Directory>


I added the top 3 lines to my access.conf file in Debian, but I can't
seem to get module imports to work for my CGIs.  What's going on?  I
don't really know Apache that well - so any help would be useful.

Vic


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



From parkw at better.net  Thu Nov 18 18:58:12 1999
From: parkw at better.net (William Park)
Date: Thu, 18 Nov 1999 18:58:12 -0500
Subject: Python in DOS
In-Reply-To: <DB6390A9770ED211B2440000F807FC7B01039B8A@nt_exlxe.lxe.com>
References: <DB6390A9770ED211B2440000F807FC7B01039B8A@nt_exlxe.lxe.com>
Message-ID: <19991118185812.A949@better.net>

On Thu, Nov 18, 1999 at 06:24:25PM -0500, Curtis, Craig M. wrote:
> Does anyone know the "approved" way to get rid of the following message:
> 
> Could not find platform dependent libraries <exec_prefix>
> Consider setting $PYTHONHOME to <prefix>[:<exec_prefix>]
> 'import site' failed; use -v for traceback
> Python 1.5.2 (#6, Nov 18 1999, 17:19:05)  [GCC 2.8.1] on ms-dos7
> Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
> 
> 
> I've tried everything I can think of to set PYTHONHOME and PYTHONPATH and
> python ALWAYS starts up this way.

If I remember (this was long long time ago), I had to remove case
sensitivity for library files by setting some dummy environment
variable.



From jcd2 at lehigh.edu  Mon Nov  8 18:58:41 1999
From: jcd2 at lehigh.edu (Josh Daghlian)
Date: Mon, 08 Nov 1999 18:58:41 -0500
Subject: [OFFTOPIC] Maps and functions (was Re: Apply a function to each list member?)
References: <slrn824gee.1f7.bwinton@tor.dhs.org> <XFMail.991105110617.mikael@isy.liu.se> <19991105215504.A497@Ridcully.home> <14370.53219.667492.946075@gargle.gargle.HOWL> <27NwjKAK1uI4EwdJ@jessikat.demon.co.uk> <m3eme59dg1.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <jcd2-0811991858420001@dial16.cc.lehigh.edu>

Here is an example of something that maps a single input to multiple outputs:

y(x) = arcsin(x)
For x = 0, y(0) == ..., -2*pi, -pi, 0, pi, 2*pi, 3*pi, ...

Note that this is *not* a function (in the math sense), but it is a
perfectly good mapping that lots of nerdy folks use all the time.

Now lets move to alt.math or something.  :)
Cheers.
--josh


In article <m3eme59dg1.fsf at atrus.jesus.cam.ac.uk>, Michael Hudson
<mwh21 at cam.ac.uk> wrote:

:  Robin Becker <robin at jessikat.demon.co.uk> writes:
:  
:  > In article <14370.53219.667492.946075 at gargle.gargle.HOWL>, Charles G
:  > Waldman <cgw at fnal.gov> writes

<cut>

:  > I define f as a thing which takes integer arguments and outputs sets of
:  > integers
:  > 
:  > presumably I can have such an f that
:  > 
:  >         f(2)={3,4}
:  > 
:  > to my simple engineer's eyes f could belong to the functions mapping Z--
:  > >2^Z
:  > 
:  > I assume we're not talking that way if the f above isn't a function.
:  
:  f is a function, as you say from Z to 2^Z. A difference between that
:  and a function from Z to Z, is that this f throws away a lot of the
:  algebraic structure of Z (i.e. being a uniquely generated commutative
:  ring), and so is in many ways less interesting than a function that
:  preserves this structure.
:  
:  Regards,
:  Michael



From news at dorb.com  Sat Nov 13 19:53:29 1999
From: news at dorb.com (Darrell)
Date: Sat, 13 Nov 1999 19:53:29 -0500
Subject: method overriding trick
References: <14378.62498.649973.150075@bitdiddle.cnri.reston.va.us> <lK9X3.696$An3.9333@newsfeed.avtel.net> <382DFF59.51F186A6@x.x>
Message-ID: <CInX3.11936$ss3.332257@typhoon.nyroc.rr.com>

Eric Jacobs <x at x.x> wrote in message news:382DFF59.51F186A6 at x.x...
> You wouldn't be able to reassign the _Parent variable to define another
> class elsewhere in the same module, though, because the functions defined
> this way will always look up the _Parent from the module namespace at
> execution time. A variant that removes that restriction is:
>
> > _Parent = basedlg.BaseDlg
> > class AddressDlg(_Parent):
> >     def __init__(self, param, _Parent=_Parent):
>
> The default args are evaluated at function definition time, which means
> that the variable has the value you expect. Now you have true cut'n'paste
> coding!

This doesn't seem to be a problem.

>>> class A:
...     a=1
...
>>> class C:
...     pass
...
>>> p=A
>>> class B(p):
...     def __init__(self, p=p):
...             print 'p:', p
...             print self.__class__.__bases__
...
>>> B().a
p: __main__.A
(<class __main__.A at 7f8f00>,)
1
>>> p=C
>>> B().a
p: __main__.A
(<class __main__.A at 7f8f00>,)
1
>>>

If you want to change base classes on the fly check this out.

>>> class X:
...     pass
...
>>> X.__bases__
()
>>> X.__bases__=(B,)
>>> X()
p: __main__.A
(<class __main__.B at 7fb760>,)
<__main__.X instance at 7fb1f0>
>>>

--
--Darrell





From Brian at digicool.com  Tue Nov 30 09:33:45 1999
From: Brian at digicool.com (Brian Lloyd)
Date: Tue, 30 Nov 1999 09:33:45 -0500
Subject: Exposing COM via XML-RPC or Something Else
Message-ID: <613145F79272D211914B0020AFF6401914DD7C@gandalf.digicool.com>

> I'm playing around with a way to expose Windows COM object via
> XML-RPC (or something else).
> 
> Let me explain:
> 
> My first thought was to write an XML-RPC API to expose the MAPI COM
> object and it's methods.
> ...
> So my current question, can anyone think of an existing technology or
> library that will allow me to expose these COM objects in a way that
> could be used by other platforms/languages. Corba (FNORB), ILU, PYRO?
> 
> Or am I really just going a little crazy?

You might want to take a look at MS' SOAP (Simple Object Access 
Protocol) spec - it's not fully baked yet, but it basically takes
xml-rpc several steps further, allowing you to communicate with 
SOAP-enabled objects (COM or otherwise) using XML over HTTP. We in 
Zope-land are keeping a close eye on this, as publishing objects 
via HTTP is the core idea of Zope already. Once the SOAP spec is 
fleshed out enough to implement in Python, it will probably be 
laugably easy not only to implement SOAP objects in Python but also
to wrap most any existing COM objects as SOAP-aware objects.

If you're interested, the current spec is at:

http://msdn.microsoft.com/workshop/xml/general/soaptemplate.asp


Brian Lloyd        brian at digicool.com
Software Engineer  540.371.6909              
Digital Creations  http://www.digicool.com 






From ionel at psy.uva.nl  Mon Nov  1 16:09:49 1999
From: ionel at psy.uva.nl (Ionel Simionescu)
Date: Mon, 1 Nov 1999 22:09:49 +0100
Subject: How to control app from Python?
References: <38160DD8.5C695E09@theriver.com> <3817BD5F.7E05BDC9@home.com> <38185A03.9AB0FAD6@theriver.com> <7vke68$6r6$1@nnrp1.deja.com>
Message-ID: <7vkvia$st7@mail.psy.uva.nl>

| > The app to be controlled is visual...
| > I discovered Mark Hammond's win32 stuff...


Easy would it be if the application is a DDE server or a COM server.
Check its docs.

If it is none of that, you can still approach it the hard way, that is by
getting a handle to one of its windows (through system functions) and then
sending messages to it, or even data (via a pipe).

The attachment is an example written in C for controlling Gnuplot.

The example is simple and clear enoough to give you a starting point.
The problem with it is that not all system functions used there are also
found in the win32 extensions, yet.

Succes,
ionel





begin 666 pgnuplot.c
M+RH*("H@<&=N=7!L;W0N8R M+2!P:7!E('-T9&EN('1O('=G;G5P;&]T"B J
M( H@*B!697)S:6]N(# N,R M+2 S,"!*=6X@,3DY.0H@*B *("H at 5&AI<R!P
M<F]G<F%M(&ES(&)A<V5D(&]N('!G;G5P;&]T+F, at 0V]P>7)I9VAT("A#*2 Q
M.3DY($AA;G,M0F5R;FAA<F0 at 0G)O96ME<@H@*B!W:71H('-U8G-T86YT:6%L
M(&UO9&EF:6-A=&EO;G, at 0V]P>7)I9VAT("A#*2 Q.3DY($-R86EG(%(N(%-C
M:&%R9'0N"B J"B J(%1H92!C;V1E(&ES(')E;&5A<V5D('1O('1H92!P=6)L
M:6, at 9&]M86EN+ at H@*@H@*B\*"B\J($-O;6UE;G1S(&9R;VT@;W)I9VEN86P@
M<&=N=7!L;W0N8R J+PHO*@H@*B!P9VYU<&QO="YC("TM("=0:7!E('1O(&=N
M=7!L;W0G("!697)S:6]N(#DY,#8P. H@*@H@*B!!('-M86QL('!R;V=R86TL
M('1O(&)E(&-O;7!I;&5D(&%S(&$@5VEN,S(@8V]N<V]L92!M;V1E(&%P<&QI
M8V%T:6]N+ at H@*B H3D(Z('=I;&P@;F]T('=O<FL@;VX@,39B:70 at 5VEN9&]W
M<RP@;F]T(&5V96X@=VET:"!7:6XS,G,@:6YS=&%L;&5D*2X*("H*("H at 5&AI
M<R!P<F]G<F%M('=I;&P at 86-C97!T(&-O;6UA;F1S(&]N(%-41$E.+"!A;F0@
M<&EP92!T:&5M('1O(&%N"B J(&%C=&EV92 H;W(@;F5W;'D at 8W)E871E9"D@
M=V=N=7!L;W0@=&5X="!W:6YD;W<N($-O;6UA;F0@;&EN92!O<'1I;VYS"B J
M(&%R92!P87-S960@;VX@=&\@=V=N=7!L;W0N"B J"B J($5F9F5C=&EV96QY
M+"!T:&ES(&UE86YS(&!P9VYU<&QO="<@:7, at 86X@86QM;W-T(&-O;7!L971E
M('-U8G-T:71U=&4*("H at 9F]R(&!W9VYU<&QO="<L(&]N('1H92!C;VUM86YD
M(&QI;F4L('=I=&@@=&AE(&%D9&5D(&)E;F5F:70@=&AA="!I= H@*B!D;V5S
M(&%C8V5P="!C;VUM86YD<R!F<F]M(')E9&ER96-T960@<W1D:6XN("A"96EN
M9R!A(%=I;F1O=W, at 1U5)"B J(&%P<&QI8V%T:6]N+"!@=V=N=7!L;W0G(&ET
M<V5L9B!C86YN;W0@<F5A9"!S=&1I;B!A="!A;&PN*0H@*@H@*B!#;W!Y<FEG
M:'0@*$,I(#$Y.3D at 8GD@2&%N<RU"97)N:&%R9"!"<F]E:V5R"B J(" @(" @
M(" @(" @(" @(" @(" @(" H8G)O96ME<D!P:'ES:6LN<G=T:"UA86-H96XN
M9&4I"B J(%1H:7, at 9FEL92!I<R!I;B!T:&4@<'5B;&EC(&1O;6%I;BX at 270@
M;6EG:'0 at 8F5C;VUE('!A<G0@;V8 at 82!F=71U<F4*("H at 9&ES=')I8G5T:6]N
M(&]F(&=N=7!L;W0N"B J"B J(&)A<V5D(&]N(&$@<')O9W)A;2!P;W-T960@
M=&\@8V]M<"YG<F%P:&EC<RYA<'!S+F=N=7!L;W0@:6X at 36%Y(#$Y.3@@8GD*
M("H@:FP at 2&%M96P@/&IL:&%M96Q 8V5A+F9R/@H@*@H@*B!#:&%N9V5S(')E
M;&%T:79E('1O('1H870@;W)I9VEN86P@=F5R<VEO;CH*("H at +2T@9&]E<VXG
M="!S=&%R="!A(&YE=R!W9VYU<&QO="!I9B!O;F4 at 86QR96%D>2!I<R!R=6YN
M:6YG+ at H@*B M+2!D;V5S;B=T(&5N9"!U<"!I;B!A;B!E;F1L97-S(&QO;W @
M:68 at 4U1$24X@:7,@;F]T(')E9&ER96-T960N"B J(" @("AR969U<V5S('1O
M(')E860 at 9G)O;2!35$1)3B!A="!A;&PL(&EN('1H870 at 8V%S92DN"B J("TM
M(&1O97-N)W0@<W1O<"!T:&4@=V=N=7!L;W0@<V5S<VEO;B!A="!T:&4 at 96YD
M(&]F"B J(" @('-T9&EN+"!I9B!I="!D:61N)W0@<W1A<G0@=V=N=7!L;W0@
M:71S96QF+ at H@*B M+2!A=F]I9',@=&AE('5S=6%L(&)U9F9E<B!O=F5R<G5N
M('!R;V)L96T@=VET:"!G971S*"DN"B J"B J($9O<B!T:&4@<F5C;W)D+"!)
M('5S=6%L;'D@=7-E($UI;D=7,S(@=&\@8V]M<&EL92!T:&ES+"!W:71H(&$*
M("H at 8V]M;6%N9"!L:6YE(&QO;VMI;F<@;&EK92!T:&ES. at H@*@H@*B @(" @
M9V-C("UO('!G;G5P;&]T+F5X92!P9VYU<&QO="YC("UL=7-E<C,R("US"B J
M+PH*+RH)36]D:69I8V%T:6]N<R!B>2!#<F%I9R!2+B!38VAA<F1T("@Q-R!*
M=6X@,3DY.2D*"0D*"4-O<'ER:6=H=" H0RD@,3DY.2!B>2!#<F%I9R!2+B!3
M8VAA<F1T("AC<F%I9T!S:6QI8V$N;7-E+G5F;"YE9'4I"@H)36%J;W(@8VAA
M;F=E<SH@*%-E92!T:&4 at 97AP;&%N871I;VX at 8F5L;W<@9F]R(&UO<F4@:6YF
M;W)M871I;VXI"@D)*R!!;'=A>7,@<W1A<G1S(&$@;F5W(&EN<W1A;F-E(&]F
M('=G;G5P;&]T+ at H)"2L at 268@<W1D:6X@:7-N)W0@<F5D:7)E8W1E9"!T:&5N
M('-T87)T('=G;G5P;&]T(&%N9"!G:79E(&ET(&9O8W5S+ at H)"2L at 57-E<R!#
M<F5A=&50<F]C97-S*"D@:6YS=&5A9"!O9B!7:6Y%>&5C*"D@=&\@<W1A<G0@
M=V=N=7!L;W0@=VAE;B!S=&1I;@H)"2 @:7,@<F5D:7)E8W1E9"X*"0H)3W1H
M97(@8VAA;F=E<SH*"0DK($YE=R!T96-H;FEQ=64 at 9F]R(&)U:6QD:6YG('1H
M92!C;VUM86YD(&QI;F4@=&\@<&%S<R!T;R!W9VYU<&QO="YE>&4*"0D@('=H
M:6-H(&ES(&QE<W, at 8V]M<&QI8V%T960 at 86YD('-E96US('1O('=O<FL@;6]R
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M+ at H*"51H92!P<F5V:6]U<R!V97)S:6]N(&]F('!G;G5P;&]T('=O=6QD(&9A
M:6P@=VAE;B!M;W)E('1H86X@;VYE('!R;V=R86T*"71R:65D('1O(&%C8V5S
M<R!W9VYU<&QO="!S:6UU;'1A;F5O=7-L>2!O<B!W:&5N(&]N92!P<F]G<F%M
M('1R:65D('1O('-T87)T"@EM;W)E('1H86X@;VYE('=G;G5P;&]T('-E<W-I
M;VXN($]N;'D at 82!S:6YG;&4@:6YS=&%N8V4@;V8@=V=N=7!L;W0@=V]U;&0@
M8F4*"7-T87)T960 at 86YD(&%L;"!I;G!U="!W;W5L9"!B92!S96YT('1O('1H
M870@:6YS=&%N8V4N(%=H96X@='=O(&]R(&UO<F4@<')O9W)A;7,*"71R:65D
M('1O('!I<&4@:6YP=70@=&\@=V=N=7!L;W0L('1H92!T=V\@<V5P97)A=&4@
M:6YP=70@<W1R96%M<R!W;W5L9"!B92!S96YT( H)=&\@;VYE('=G;G5P;&]T
M('=I;F1O=R!R97-U;'1I;F<@:6X@;VYE('9E<GD at 8V]N9G5S960 at 8V]P>2!O
M9B!W9VYU<&QO="X at 5&AE(&]N;'D*"7=A>2!T;R!A=F]I9"!T:&ES('!R;V)L
M96T@=V%S('1O(&-H86YG92!P9VYU<&QO="!S;R!T:&%T(&ET('=O=6QD('-T
M87)T(&$@"@EN97<@:6YS=&%N8V4@;V8@=V=N=7!L;W0 at 979E<GD@=&EM92X*
M"@E*=7-T('-T87)T:6YG(&$@;F5W(&EN<W1A;F-E(&]F('=G;G5P;&]T(&ES
M;B=T(&5N;W5G:"X@<&=N=7!L;W0@;75S="!A;'-O"@EM86ME('-U<F4@=&AA
M="!T:&4 at 9&%T82!O;B!E86-H('-T9&EN('!I<&4@:7,@<V5N="!T;R!T:&4@
M<')O<&5R('=G;G5P;&]T( H):6YS=&%N8V4N(%1H:7,@:7, at 86-H:65V960@
M8GD@=7-I;F<@0W)E871E4')O8V5S<R at I('=H:6-H(')E='5R;G, at 82!H86YD
M;&4@"@ET;R!T:&4@;F5W;'D at 8W)E871E9"!P<F]C97-S+B!/;F-E('1H92!P
M<F]C97-S(&AA<R!I;FET:6%L:7IE9"P@:70 at 8V%N(&)E( H)<V5A<F-H960@
M9F]R('1H92!T97AT('=I;F1O=R!A;F0@=&AE;B!D871A(&-A;B!B92!R;W5T
M960 at 8V]R<F5C=&QY+B!4:&4@<V5A<F-H( H):7, at 8V%R<FEE9"!O=70 at 8GD@
M=&AE($5N=6U4:')E8617:6YD;W=S*"D at 8V%L;"!A;F0@=&AE(&1A=&$@<&%S
M<VEN9R!I<R!C87)R:65D( H);W5T(&)Y(&$@<F5W<FET=&5N('9E<G-I;VX@
M;V8@=&AE(&]R:6=I;F%L(&-O9&4N(%=I=&@@=&AE<V4 at 8VAA;F=E<RP@<&=N
M=7!L;W0@"@EN;W<@8F5H879E<R!I;B!A(&UA;FYE<B!C;VYS:7-T96YT('=I
M=&@@=&AE(&)E:&%V:6]R(&]F(&=N=7!L;W0@;VX at 54Y)6" *"6-O;7!U=&5R
M<RX*"@E4:&ES('!R;V=R86T@:&%S(&)E96X at 8V]M<&EL960@=7-I;F<@36EC
M<F]S;V9T(%9I<W5A;"!#*RL at -"XP('=I=&@@=&AE"@EF;VQL;W=I;F<@8V]M
M;6%N9"!L:6YE. at H*"0EC;" O3S(@<&=N=7!L;W0N8R O;&EN:R!U<V5R,S(N
M;&EB"@H)5&AE(')E<W5L=&EN9R!P<F]G<F%M(&AA<R!B965N('1E<W1E9"!O
M;B!7:6Y.5"!A;F0 at 5VEN.3@@8F]T:"!B>2!C86QL:6YG"@EI="!D:7)E8W1L
M>2!F<F]M('1H92!C;VUM86YD(&QI;F4@=VET:"!A;F0@=VET:&]U="!R961I
M<F5C=&5D(&EN<'5T+B!4:&4*"7!R;V=R86T at 86QS;R!W;W)K<R!O;B!7:6Y.
M5"!W:71H(&$@;6]D:69I960@=F5R<VEO;B!O9B!';G5P;&]T+G!Y("AA('-C
M<FEP= H)9F]R(&EN=&5R86-T:79E(&-O;G1R;VP@;V8 at 1VYU<&QO="!F<F]M
M(%!Y=&AO;BDN( H)"@DR,B!*54X@,3DY.3H*"2L at 1FEX960@8V]M;6%N9"!L
M:6YE(&-O9&4@=&\@8F5H879E('!R;W!E<FQY('=H96X@=&AE(&9I<G-T"@D@
M(&ET96T@:7,@<75O=&5D(&EN('1H92!O<FEG:6YA;"!C;VUM86YD(&QI;F4N
M"@D*"3(Y($I53B Q.3DY. at H)*R!!9&1E9"!S;VUE(&-O9&4@=&\@<')I;G0@
M=&AE(&-O;6UA;F0@;&EN92X at 5&AI<R!I<R!F;W(@=&5S=&EN9PH)("!O;FQY
M(&%N9"!S:&]U;&0 at 8F4@<F5M;W9E9"!B969O<F4@=&AE(&=E;F5R86P@<F5L
M96%S92X at 5&\@96YA8FQE+" *"2 @8V]M<&EL92!W:71H(%-(3U=#341,24Y%
M(&1E9FEN960N"@H),S @2E5.(#$Y.3DZ"@DK($YE=R!F=6YC=&EO;B!&:6YD
M56YQ=6]T9613<&%C92 at I('=H:6-H(')E<&QA8V5S('1H92!E87)L:65R('1E
M8VAN:7%U92!F;W(*"2 @9FEN9&EN9R!T:&4 at 8V]M;6%N9"!L:6YE(&%R9W5M
M96YT<R!T;R!S96YD(&]N('1O('=G;G5P;&]T+B!0<FEO<B!T;R!T:&ES"@D@
M('1H92!A<F=U;65N=',@=V5R92!A<W-U;65D('1O('-T87)T(&%F=&5R(&%R
M9W9;,%TL(&AO=V5V97(@87)G=ELP72!I<R!N;W0*"2 @<V5T('1H92!S86UE
M(&)Y(&%L;"!C;VUB:6YI=&%I;VYS(&]F(&-O;7!I;&5R+"!C;VUM86YD('!R
M;V-E<W-O<BP at 86YD($]3+B *"2 @5&AE(&YE=R!M971H;V0@:6=N;W)E<R!A
M<F=V(&-O;7!L971E;'D at 86YD(&UA;G5A;&QY('-E87)C:"!T:&4 at 8V]M;6%N
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M(")W9VYU<&QO="YE>&4B"B\J($-24SH at 5&AE('9A;'5E(&=I=F5N(&%B;W9E
M('=I;&P@=V]R:R!C;W)R96-T;'D at 87,@;&]N9R!A<R!P9VYU<&QO="YE>&4*
M("H@:7,@:6X@=&AE('-A;64 at 9&ER96-T;W)Y(&%S('=G;G5P;&]T+F5X92!O
M<B!T:&4 at 9&ER96-T;W)Y(&-O;G1A:6YI;F<*("H@=V=N=7!L;W0N97AE(&ES
M(&EN8VQU9&5D(&EN('1H92!P871H+B!)('=O=6QD(')E8V]M;65N9"!P;&%C
M:6YG('1H92 *("H@<&=N=7!L;W0N97AE(&5X96-U=&%B;&4@:6X@=&AE('-A
M;64 at 9&ER96-T;W)Y(&%S('=G;G5P;&]T+F5X92!A;F0@"B J(&QE879I;F<@
M=&AI<R!D969I;FET:6]N(&%L;VYE+ at H@*B\*"B-D969I;F4 at 5TE.1$]73D%-
M12 B9VYU<&QO="(*(V1E9FEN92!005)%3E1#3$%34R B=V=N=7!L;W1?<&%R
M96YT(@HC9&5F:6YE(%1%6%1#3$%34R B=V=N=7!L;W1?=&5X="(*(V1E9FEN
M92!'4D%02%=)3D1/5R B9VYU<&QO="!G<F%P:"(*(V1E9FEN92!'4D%02$-,
M05-3(")W9VYU<&QO=%]G<F%P:"(*(V1E9FEN92!"549&15)?4TE:12 X, H*
M+RH at 1TQ/0D%,(%9A<FEA8FQE<R J+PI(5TY$(&AW;F1087)E;G0@/2!.54Q,
M.PI(5TY$(&AW;F1497AT(#T at 3E5,3#L*"E!23T-%4U-?24Y&3U)-051)3TX@
M<&E0<F]C26YF;SL@"E-405)455!)3D9/(" @(" @(" @<VE3=&%R=$EN9F\[
M( H*+RH at 0U)3.B!#86QL8F%C:R!F;W(@=&AE($5N=6U4:')E8617:6YD;W=S
M(&9U;F-T:6]N("HO"D)/3TP at 0T%,3$)!0TL at 8V)'971497AT5VEN9&]W*$A7
M3D0@(&AW;F0L($Q005)!32 @;%!A<F%M("D*>PH)+RH@<V%V92!T:&4@=F%L
M=64@;V8@=&AE('!A<F5N="!W:6YD;W<@*B\*"6AW;F1087)E;G0@/2!H=VYD
M.PH)+RH at 8VAE8VL@=&\@<V5E(&EF(&ET(&AA<R!A(&-H:6QD('1E>'0@=VEN
M9&]W("HO"@EH=VYD5&5X=" ]($9I;F17:6YD;W=%>"AH=VYD+"!.54Q,+"!4
M15A40TQ!4U,L($Y53$P@*3L*"@DO*B!I9B!T:&4@=&5X="!W:6YD;W<@=V%S
M(&9O=6YD+"!S=&]P(&QO;VMI;F<@*B\*"7)E='5R;B H(&AW;F1497AT("$]
M($Y53$P@*3L*?0H*+RH@<V5N9', at 82!S=')I;F<@=&\@=&AE('-P96-I9FEE
M9"!W:6YD;W<@*B\*+RH at 0U)3.B!M861E('1H:7,@:6YT;R!A(&9U;F-T:6]N
M(&-A;&P@*B\*=F]I9"!0;W-T4W1R:6YG*$A73D0@:'=N9"P at 8VAA<B J<&,I
M"GL*"7=H:6QE*" J<&,@*7L*"0E0;W-T365S<V%G92@@:'=N9"P at 5TU?0TA!
M4BP@*G!C+" Q3" I.PH)"2\J($-24SH@<VAO=6QD(&%D9"!A(&-H96-K(&]F
M(')E='5R;B!C;V1E(&]N(%!O<W1-97-S86=E+B!)9B P+"!T:&4@"@D)(" @
M;65S<V%G92!Q=64@=V%S(&9U;&P at 86YD('1H92!M97-S86=E('=A<VXG="!P
M;W-T960N("HO"@D)<&,K*SL*"7T*?0H*+RH at 1FEN9%5N<75O=&5D4W!A8V4H
M*3H at 4V5A<F-H(&$@<W1R:6YG(&9O<B!T:&4 at 9FER<W0@<W!A8V4@;F]T(&5N
M8VQO<V5D(&EN('%U;W1E<RX*("H@("!2971U<FYS(&$@<&]I;G1E<B!T;R!T
M:&4@<W!A8V4L(&]R('1H92!E;7!T>2!S=')I;F<@:68@;F\@<W!A8V4@:7,@
M9F]U;F0N( H@*B @("U#4E,@,S P-C$Y.3D@( H@*B\*8VAA<BH at 1FEN9%5N
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M<'-#;61,:6YE(#T at 3E5,3#L*"4)/3TP@(" @8E-U8V-E<W,@/2!&04Q313L*
M"0H)+RH at 2$)"(#$Y.3DP,S(U+"!T;R!A;&QO=R!P9VYU<&QO=" \('-C<FEP
M=" ^(&]U='!U="YG:68@*B\*"5]S971M;V1E*&9I;&5N;RAS=&1O=70I+"!?
M3U]"24Y!4EDI.PH*"2\J($-24SH at 8W)E871E('1H92!N97<@8V]M;6%N9"!L
M:6YE+"!P87-S:6YG(&%L;"!O9B!T:&4 at 8V]M;6%N9 H)("H@;&EN92!O<'1I
M;VYS('1O('=G;G5P;&]T('-O('1H870@:70 at 8V%N('!R;V-E<W,@=&AE;3H*
M"2 J(&9I<G-T+"!G970@=&AE(&-O;6UA;F0@;&EN92P*"2 J('1H96X@;6]V
M92!P87-T('1H92!N86UE(&]F('1H92!P<F]G<F%M("AE+F<N+" G<&=N=7!L
M;W0G*2P*"2 J(&9I;F%L;'DL(&%D9"!W:&%T)W,@;&5F="!O9B!T:&4@;&EN
M92!O;G1O('1H92!G;G5P;&]T(&-O;6UA;F0@;&EN92X@*B\*"7!S0VUD3&EN
M92 ]($=E=$-O;6UA;F1,:6YE*"D["@HC:69D968 at 4TA/5T--1$Q)3D4*"69P
M<FEN=&8H<W1D97)R+")#;61,:6YE.B E<UQN(BP@<'-#;61,:6YE*3L*"69P
M<FEN=&8H<W1D97)R+")A<F=V6S!=.B E<UQN(BQA<F=V6S!=*3L*(V5N9&EF
M"@H)+RH at 0U)3(#,P,#8Q.3DY.B!396%R8V@@9F]R('1H92!F:7)S="!U;G%U
M;W1E9"!S<&%C92X at 5&AI<R!S:&]U;&0@"@D@("!S97!A<F%T92!T:&4@<')O
M9W)A;2!N86UE(&9R;VT@=&AE(&%R9W5M96YT<RX@*B\*"7!S0VUD3&EN92 ]
M($9I;F15;G%U;W1E9%-P86-E*"!P<T-M9$QI;F4@*3L*"@ES=')N8V%T*'!S
M1VYU<&QO=$-O;6UA;F1,:6YE+"!P<T-M9$QI;F4L($U!6%]0051(("T@<W1R
M;&5N*'!S1VYU<&QO=$-O;6UA;F1,:6YE*2D["@HC:69D968 at 4TA/5T--1$Q)
M3D4*"69P<FEN=&8H<W1D97)R+")!<F=U;65N=',Z("5S7&XB+"!P<T-M9$QI
M;F4I.PH)9G!R:6YT9BAS=&1E<G(L(D=N=7!L;W1#;VUM86YD3&EN93H@)7-<
M;B(L<'-';G5P;&]T0V]M;6%N9$QI;F4I.PHC96YD:68*"@DO*B!#4E,Z(&EF
M('-T9&EN(&ES;B=T(')E9&ER96-T960@=&AE;B!J=7-T(&QA=6YC:"!W9VYU
M<&QO="!N;W)M86QL>0H)("H at 86YD(&5X:70N("HO"@EI9B H(&ES871T>2AF
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M3%5212D["@E]"@H)+RH at 0U)3.B!I;FET:6%L:7IE('1H92!35$%25%5024Y&
M3R!A;F0 at 8V%L;"!#<F5A=&50<F]C97-S*"DN("HO( H)<VE3=&%R=$EN9F\N
M8V(@/2!S:7IE;V8H4U1!4E154$E.1D\I.R *"7-I4W1A<G1);F9O+FQP4F5S
M97)V960@/2!.54Q,.R *"7-I4W1A<G1);F9O+FQP4F5S97)V960R(#T at 3E5,
M3#L@"@ES:5-T87)T26YF;RYC8E)E<V5R=F5D,B ](# [( H)<VE3=&%R=$EN
M9F\N;'!$97-K=&]P(#T at 3E5,3#L@"@ES:5-T87)T26YF;RYD=T9L86=S(#T@
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M<V5C=7)I='D at 871T<FEB=71E<R @*B\*"0D)1D%,4T4L(" @(" @(" @(" @
M(" @(" @+RH@:&%N9&QE(&EN:&5R:71A;F-E(&9L86<@(" @(" @(" @(" @
M(" @*B\*"0D),"P@(" @(" @(" @(" @(" @(" @(" @+RH at 8W)E871I;VX@
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M(" @*B\*"0D)*3L*"@DO*B!I9B!#<F5A=&50<F]C97-S*"D at 9F%I;&5D+"!P
M<FEN="!A('=A<FYI;F<@86YD(&5X:70N("HO"@EI9B H("$@8E-U8V-E<W,@
M*2!["@D)9G!R:6YT9BAS=&1E<G(L(D524D]2("5U.B!#;W5L9&XG="!E>&5C
M=71E.B!<(B5S7")<;B(L( H)"0D)"0E'971,87-T17)R;W(H*2P@<'-';G5P
M;&]T0V]M;6%N9$QI;F4I.PH)"65X:70H15A)5%]&04E,55)%*3L*"7T*"@DO
M*B!#4E,Z(&=I=F4 at 9VYU<&QO="!E;F]U9V@@=&EM92!T;R!S=&%R=" H,2!S
M96,N*2 J+PH):68@*"!786ET1F]R26YP=71)9&QE*'!I4')O8TEN9F\N:%!R
M;V-E<W,L(#$P,# I("D@>PH)"69P<FEN=&8H<W1D97)R+" B5&EM96]U=#H@
M9VYU<&QO="!I<R!N;W0@<F5A9'E<;B(I.PH)"65X:70H15A)5%]&04E,55)%
M*3L*"7T*"@DO*B!#4E,Z(&=E="!T:&4 at 2%=.1"!O9B!T:&4@<&%R96YT('=I
M;F1O=R!A;F0@=&5X="!W:6YD;W=S("HO"@E%;G5M5&AR96%D5VEN9&]W<RAP
M:5!R;V-);F9O+F1W5&AR96%D260L(&-B1V5T5&5X=%=I;F1O=RP@,"D["@H)
M+RH at 0U)3.B!F<F5E('1H92!P<F]C97-S(&%N9"!T:')E860@:&%N9&QE<R J
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M?0H*"2\J('=A:70 at 9F]R(&-O;6UA;F1S(&]N('-T9&EN+"!A;F0@<&%S<R!T
M:&5M(&]N('1O('1H92!W9VYU<&QO="!T97AT"@D@*B!W:6YD;W<@*B\*"7=H
M:6QE("@@9F=E=',H<'-"=69F97(L($)51D9%4E]325I%+"!S=&1I;BD@(3T@
M3E5,3" I('L*"0E0;W-T4W1R:6YG*&AW;F1497AT+"!P<T)U9F9E<BD["@E]
M"@D*"2\J(&5X:70 at 9W)A8V5F=6QL>2 J+PH)+RH at 0U)3.B!!9&0 at 82!T97-T
M('1O('-E92!I9B!G;G5P;&]T(&ES('-T:6QL(')U;FYI;F<_("HO"@E0;W-T
M4W1R:6YG*&AW;F1497AT+" B7&YE>&ET7&XB*3L*"@ER971U<FX at 15A)5%]3
*54-#15-3.PI]"@``
`
end




From pduffin at mailserver.hursley.ibm.com  Mon Nov  8 06:05:58 1999
From: pduffin at mailserver.hursley.ibm.com (Paul Duffin)
Date: Mon, 08 Nov 1999 11:05:58 +0000
Subject: embedding stubbed tcl
References: <bLd7vSA1euH4EwDS@jessikat.demon.co.uk>
	 <8ZCT3.18814$1v5.202282@newscontent-01.sprint.ca>
	 <VVo0SGAJExH4EwyM@jessikat.demon.co.uk> <3821E345.FC6C372F@chello.nl> <Ik5jOCAzYgI4Ewad@jessikat.demon.co.uk>
Message-ID: <3826AE96.F36@mailserver.hursley.ibm.com>

Robin Becker wrote:
> 
> In article <3821E345.FC6C372F at chello.nl>, Jan Nijtmans
> <j.nijtmans at chello.nl> writes
> >Robin Becker wrote:
> >
> >> I thought stubs were for extensions. The tkinter case is slightly the
> >> wrong way round as it starts tcl up.
> >
> >Stubs can very well be used for applications as well. For example,
> >I already used it to Stubify the Tcl plugin, so I don't need to
> >recompile it when Tcl8.3 comes out ;-).  The only problem is that
> >Tcl doesn't yet contain a portable way to use dlopen()/LoadLibrary()
> >or whatever this function might be called.
> >
> ok so what is the minimal way to do this?

Does tkinter call Tk_Main (doubtful) ?
I would suggest that you do something like.

	hTcl = dlopen (whatever version of Tcl)
	createInterp = dlsym (hTcl, "Tcl_CreateInterp");
	interp = createInterp ();
	Tcl_InitStubs (interp, "version", ...);

If Tk was a dynamically loadable package then you could do something
like

	Tcl_PkgRequire (interp, "Tk", "version", ...);

Otherwise you have to do

	hTk = dlopen (whatever version of Tk)
	tkInit = dlsym (hTk, "Tk_Init")
	tkSafeInit = dlsym (hTk, "Tk_SafeInit")
	tkInit (interp);
	Tcl_StaticPackage (tkInit, tkSafeInit, "Tk", "version")

At this point you now have a working Tcl and Tk. How you actually choose
the version of Tcl and Tk to use is up to you.

-- 
Paul Duffin
DT/6000 Development	Email: pduffin at hursley.ibm.com
IBM UK Laboratories Ltd., Hursley Park nr. Winchester
Internal: 7-246880	International: +44 1962-816880



From parkw at better.net  Wed Nov 10 06:31:34 1999
From: parkw at better.net (William Park)
Date: Wed, 10 Nov 1999 06:31:34 -0500
Subject: Text Manipulation in Python
In-Reply-To: <VA.00000012.055a8702@ops.corpex.com>
References: <VA.00000012.055a8702@ops.corpex.com>
Message-ID: <19991110063134.A3633@better.net>

On Wed, Nov 10, 1999 at 08:46:42AM +0000, Edward Hasted wrote:
> I am completely new to Python.
> 
> We want to use it to alter specific lines in template files, typically 
> something like changing:-
> 
> Variable = 1234
> 
> to 
> 
> Variable = Company Name
> 
> The text manipulation strings within Python seem to work sequentially 
> rather than on a line basis.
> 
> 1. Is this correct?

Well, you have read line by line.  Try something like
    f = open(..., "r")
    for s in f.readlines():
	s = ...do string substitution using 're' or 'string' module...
	print s
But, if you're running Unix, you can also use 'sed'.

> 2. What is the best way to do text searching and manipulation within 
> Python.
> 
> Many thanks,
> 
> Edward 
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list



From emuller at painewebber.com  Tue Nov  9 10:23:51 1999
From: emuller at painewebber.com (emuller at painewebber.com)
Date: Tue, 09 Nov 1999 15:23:51 GMT
Subject: Running a program as another user on Win32
Message-ID: <809ea4$ln$1@nnrp1.deja.com>

I am writing a bunch of scripts that will need to run as Administrator
(Domain Administrator) on Windows NT 4.0 workstation clients that
belong to an NT domain. I've tried using LogonUser, but I always get
the following: pywintypes.api_error: (1314, 'LogonUser', ' A required
privilege is not held by the client.'). I've also tried using LogonUser
with the local Administrator account on my NT machine with no luck.

The scripts will be run locally on the system and perform various
tasks. The scripts will be used by service technicians and admins. If I
can't figure out how to do this in Python I'll have to use WinBatch,
which supposedly can do this (Run programs as another user)...

Any ideas how I can do this? Help! I don't want to use winbatch.


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



From fbi at boxer.no  Tue Nov  9 08:47:49 1999
From: fbi at boxer.no (Frank Biringer)
Date: Tue, 09 Nov 1999 14:47:49 +0100
Subject: Writing dates to a ODBC-database
Message-ID: <38282605.3CD233E9@boxer.no>

Does anybody know how to write a date to a ODBC-database?

I use the standard dbi- and odbc-modules for the Python-win32
distribution. I guess I have to create a dbi.dbiDate object. But since
there's neither source or documentation for the dbi-module I don't know
what kind of parameter this method expects (it expects exactly one
parameter).

Any ideas??

--
Regards,
_________
 -|- rank Biringer.

______________________________________________________
Systemanalytiker                 Boxer Technologies AS
tel: (+47) 55 19 20 34           mailto:fbi at boxer.no
fax: (+47) 38 01 90 01           http://www.boxer.no/
______________________________________________________





From a_stepanov at yahoo.com  Sat Nov  6 22:52:37 1999
From: a_stepanov at yahoo.com (Alexander Stepanov)
Date: Sat, 06 Nov 1999 22:52:37 -0500
Subject: Python and generating images for web
Message-ID: <3824F785.40DF4393@yahoo.com>

Hello,


I am a newcomer to web programming and python.  I need to create a small
web application
to make some charts based on user's data input.

Is it possible to generate gif or png images on a fly using Python?

I will appreciate if you point me out accessible chart/plot  modules and
libraries as well as documentation and
working sites.    Any additional information is strongly welcome.

(I am in a process of comparison Perl and Python to solve this problem.
I plan to use the application
under WindowsNT4.0 environment + Perl? or Python interp-r).

Sincerely,  Alexander
email: a_stepanov at yahoo.com





From arad at fc.hp.com  Sat Nov  6 22:29:07 1999
From: arad at fc.hp.com (Arad Rostampour)
Date: Sat, 6 Nov 1999 20:29:07 -0700
Subject: win32com and IIS?
Message-ID: <802s6a$mj5$1@web1.cup.hp.com>

Hi,

I've been using win32com to access IE and get webpages using:

import win32com.client
ie = win32com.client.Dispatch("InternetExplorer.Application")
ie.Navigate("http://www.hp.com")
...

This works fine from the command line.  But, when I try to do this from a
CGI script under IIS, I get the following error:

File "f:\python\win32com\client\dynamic.py", line 319, in __getattr__ raise
pythoncom.com_error, details pywintypes.com_error: (-2147467259,
'Unspecified error', (0, None, None, None, 0, -2147467259), None) Exception
exceptions.AttributeError: in ignored

Any ideas on how to work around this?

Thanks,
Arad
--
Arad Rostampour- arad at fc.hp.com
NT Platform and I/O Team
Software and System Development Lab
Hewlett-Packard





From jam at quark.emich.edu  Wed Nov 24 17:31:28 1999
From: jam at quark.emich.edu (Jeff)
Date: Wed, 24 Nov 1999 17:31:28 -0500
Subject: newbie question
In-Reply-To: <943469319.838266356@news.teleweb.at>
References: <943469319.838266356@news.teleweb.at>
Message-ID: <19991124173128.A26194@quark.emich.edu>

On Wed, Nov 24, 1999 at 06:57:02PM +0000, Markus Krutz wrote:
> How the f*** do I create a string containing a *single* "\" ?
> 
> regards
> Markus
> 

[~] [5:30pm] [jam at toast-pts/3] % python
Python 1.5.2 (#1, Apr 18 1999, 16:03:16)  [GCC pgcc-2.91.60 19981201
(egcs-1.1.1  on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> a = "\\"
>>> print a
\

piece-of-pie-ly y'rs,
J
-- 
|| visit gfd <http://quark.emich.edu/>
|| psa member #293 <http://www.python.org/> 
|| New Image Systems & Services, Inc. <http://www.newimage.com/>



From tbryan at python.net  Thu Nov 11 21:12:09 1999
From: tbryan at python.net (Thomas A. Bryan)
Date: Thu, 11 Nov 1999 21:12:09 -0500
Subject: ASCII delimited files
References: <942231502.1914033961@news.interbusiness.it> <382A240B.B247AD6B@python.net> <382A2801.FDE3CA09@python.net> <382A4330.833C4EBD@easystreet.com>
Message-ID: <382B7779.9601F556@python.net>

Al Christians wrote:
> 
> There's a trick with files of delimited fields.  If a string field
> contains a delimiter (e.g. a comma in a csv file), then the field
> gets enclosed in quotes.  

Ah.  That's probably what the poster wanted!  I just posted what I had 
lying around on my machine.  In my case, I had to deal with basically 
those three field types, numeric values in a range, values from a list, 
and values that matched a regex (like a date).  They never contained 
the delimeter (usually a tab or the pipe '|' character).  

I guess I should have said that it helped me, but YMMV. :)

The big problem for me was that the values came from hardware that 
sometimes garbled dates or did other nasty things.  The format of the 
file (number and ordering of the columns) wasn't set in stone either, 
so I needed the capability to detect whether a new data file given 
to me possibly deviated from the one the parser expected.  Thus, the 
ability to build a parser quickly that would validate the data was 
what I wanted.  I still like my idea since, in my opinion, one of the 
worst things is loading corrupt data into a database. The parser code 
should deal with delimiter characters nested in the values.  I don't 
need it, and I lack sufficient leisure time to do it just for fun right 
now.  If somebody else has the need/time, feel free to take the code 
and make something really usable out of it.



---Tom

> What I usually wind up doing to circumvent these problems with delimited
> files is loading them back into a spreadsheet, typically the same one
> that wrote them (I suppose that a database program would
> offer the same option), and then I write them out as tab-delimited.
> Tabs within fields are just about never in a lifetime in my work, so
> far.



From tim_one at email.msn.com  Tue Nov 30 01:59:45 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Tue, 30 Nov 1999 01:59:45 -0500
Subject: performance of tuple-less Python?
In-Reply-To: <HSx04.8911$cz3.321455@nnrp1.uunet.ca>
Message-ID: <000301bf3b00$7c775f60$c92d153f@tim>

[Greg Wilson]
> Has anyone ever built a version of Python that didn't include tuples,
> and looked at the performance of code that consistently used lists
> instead?  I'm curious as to whether the use of tuples (which I believe
> are conceptually redundant)

They do differ semantically, due to immutability (see other msgs).

> actually has a significant positive impact on the interpreter's
> performance on non-trivial codes (i.e. once caching, instruction
> ordering, and all other effects have come into play).

tuple[i] and list[i] access have (up to accidents of architecture) identical
cost; tuple[i] may be faster in pathological cases because of cache effects
(a tuple's object header is contiguous with the tuple's content vector; a
list's object header is not generally contiguous with the list's content
vector).

A predictably bigger effect is that allocating a list object requires two
malloc() calls (one for the object header, another for the content vector).
Allocating a tuple object typically requires no malloc calls (because tuples
are expected to be short, Python maintains internal free lists of tuples of
various small sizes, and recycles them directly from there).  Similarly,
destroying a list requires two free() calls, while destroying a tuple
typically just links it onto the appropriate internal free list.

So creation and destruction of tuples goes faster, and tuples are more
memory-efficient.  Indeed, I've had larger programs that wouldn't run at all
on a 32Mb machine (constant disk thrashing) before I changed lists to
tuples!

That doesn't mean all possible implementations would have these
characteristics; they're the truth for the implementation that exists,
though.

if-tuples-didn't-exist-someone-would-invent-them-ly y'rs  - tim





From paul at digicool.com  Mon Nov  8 07:48:31 1999
From: paul at digicool.com (Paul Everitt)
Date: Mon, 08 Nov 1999 07:48:31 -0500
Subject: Production Web Deployment?
References: <NBBBKMMANMGEEBHPJJECGEGIFJAA.dirkl@home.com>
Message-ID: <3826C69F.B4E26B89@digicool.com>

Dirk Leas wrote:
> I've read numerous threads regarding how to deploy Python on the web and
> there appears to be no real consensus on what actually performs/scales well
> (about 3M hits/month). Is anybody taking lots of hits with httpdapy,
> PyApache, ASP+Python, fastcgi, or Zope? I've got a customer who's
> development staff is struggling with the complexity of Java Servlets and
> they're already using a large 3rd party, vertical Python application (and
> several folks already use Python). Peliminary testing showed plain CGI was
> just too slow...

Zope.org seems to be doing up to 2M hits/month quite easily, and it's a
pretty complicated site (many of the pages are generated by catalog
queries to the index).  Over the last couple of weeks we had some
inexplicable slow spots, which turned out to be caused by accidentally
setting the object cache way too low (200 instead of 4000).  Note that
Zope.org doesn't use a relational database for any content.

As far as I can tell, performance on Zope.org is quite snappy right
now.  It's running on a sub-$1k machine (a vanilla PIII-550 with 256 Mb
RAM and an IDE drive).  I'd expect it to do four or five times more hits
before it appears sluggish.

As mentioned in another reply, keeping database connections open is
quite important.  Zope does a good job at this and some of the database
adapters are pretty well optimized (Oracle and Sybase particularly).

Hopefully over the next month we'll migrate Zope.org away from PCGI (a
Zope system for getting requests from the web server to the long-running
process) to FastCGI.  This will eliminate execution of small C "wrapper"
executable.

Finally, Digital Creations has a commercial-with-source-plus-consulting
solution called ZEO:

  http://www.digicool.com/Solutions/ZEOFactSheet.pdf

This lets you spread your object system across multiple
machines/locations.

--Paul



From parkw at better.net  Sun Nov 14 07:18:33 1999
From: parkw at better.net (William Park)
Date: Sun, 14 Nov 1999 07:18:33 -0500
Subject: Python and sending emails with it
In-Reply-To: <382E2DB2.A00B9109@nv2.cc.va.us>
References: <382E2DB2.A00B9109@nv2.cc.va.us>
Message-ID: <19991114071833.B2775@better.net>

On Sat, Nov 13, 1999 at 10:34:10PM -0500, Jeremy Misavage wrote:
> I'm kinda new to python and I'm trying to write a small program that will take
> in form data, save it in a db, and then send an email to a person.
> 
> The program processes a users request for a "equipment request."  It gets
> loggged in a db (already done) and sends an email to the person in change of the
> equipment.  I need help on the email part.  Does some out there have some code
> that will taken data and send it to a person.  It shouldn't need a username and
> password, but if it does I'll live with that.

In Python, try 'smtplib' module.  I think it goes something like
    e = smtplib.SMTP('machine.com')
    e.sendmail('tekhir', 'supervisor', "...")
    e.quit()
Or, if you use Linux machine (like me), pipe the message to 
    mail -s "subject" supervisor

	Yours truly,
	William Park



From skip at mojam.com  Mon Nov 29 15:59:01 1999
From: skip at mojam.com (Skip Montanaro)
Date: Mon, 29 Nov 1999 14:59:01 -0600 (CST)
Subject: Python bugs
In-Reply-To: <51C9F3C392D0D111A15600A0C99E64A9AF5690@localhost>
References: <51C9F3C392D0D111A15600A0C99E64A9AF5690@localhost>
Message-ID: <14402.59669.289009.979094@dolphin.mojam.com>

    Python 1.5.2 (#0, Apr 13 1999, 10:51:12) [MSC 32 bit (Intel)] on win32
    Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
    >>> import os
    >>> {}.update(os.environ)
    Traceback (innermost last):
      File "<stdin>", line 1, in ?
    TypeError: dictionary, instance
    >>> import shelve
    >>> s = shelve.open('foo')
    >>> s.get('bar', 'baz')
    Traceback (innermost last):
      File "<stdin>", line 1, in ?
    AttributeError: get
    >>> 

I believe os.environ is a class to allow constructs like

    os.environ["FOO"] = "PYTHON"

to work like

    os.putenv("FOO", "PYTHON")

with regard to setting or changing the values of environment variables.

The lack of a get attribute in shelve objects seems to have been rectified
since the 1.5.2 release:

    >>> db = shelve.BsdDbShelf (bsddb.btopen("/var/tmp/foo2", "c"))
    >>> db["1"] = 1
    >>> db.keys()
    ['1']
    >>> db.get("1", 0)
    1
    >>> db.get("2", 0)
    0                        

Skip Montanaro | http://www.mojam.com/
skip at mojam.com | http://www.musi-cal.com/
847-971-7098   | Python: Programming the way Guido indented...




From fredrik at pythonware.com  Wed Nov 10 07:03:58 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 10 Nov 1999 13:03:58 +0100
Subject: Python company worth $100M
References: <38295BC2.13E2E31F@digicool.com>
Message-ID: <033201bf2b73$ad1e49e0$f29b12c2@secret.pythonware.com>

Paul Everitt <paul at digicool.com> wrote:
> Nice guys _don't_ finish last, it seems.  Congrats to Scott and Sam, and
> yippee for us, because this is good news!

here's another interesting story:
http://cnbc.com/commentary/commentary_full_story_stocks.asp?StoryID=6926

    "Searchbutton.com Inc. is a rare breed of Silicon Valley
    start-up. The CEO of the nine-month-old firm says his
    company will actually be making money before it goes
    public, which could happen as early as next year."

    "I'd certainly buy the stock when they go public," says
    Darcy Fowkes, an analyst with the Aberdeen Group /.../
    They have exactly the things Wall Street looks for."

and if you look at the technical credits page, what do you
find?

    http://www.searchbutton.com/misc/technicalCredits

asynclib again, of course ;-)

</F>




From rm at murl.com  Tue Nov  9 09:54:25 1999
From: rm at murl.com (Robert Meegan)
Date: Tue, 09 Nov 1999 14:54:25 GMT
Subject: [Python] header (fwd)
In-Reply-To: <Pine.HPX.4.20.9911081711570.27637-100000@ztt.rsn.hp.com>
References: <Pine.HPX.4.20.9911081711570.27637-100000@ztt.rsn.hp.com>
Message-ID: <19991109.14542500@robert.polytopic.com>

>>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<<

On 11/8/99, 5:13:19 PM, Benoit Ciceron <bciceron at rsn.hp.com> wrote 
regarding [Python] header (fwd):


> can we get an automatic subject header : "[Python]" to ALL messages
> submitted to the list , including replies and newsgroup threads ?

> thank you,

> ben_

If what you want is a way to filter mailing list messages, may I 
suggest that you look for "python-list" in both the "To:" and "CC:" 
fields. If you would like to sort your mail, "[Python]" is not 
adequate, as you will need to add the capability to move all of 
Guido's and Tim Peters' mail to the top while moving anything with 
"Perl" in the subject or body to the bottom.

--- Robert


> --
> http://www.python.org/mailman/listinfo/python-list






From fredrik at pythonware.com  Wed Nov 10 05:51:00 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 10 Nov 1999 11:51:00 +0100
Subject: [Python-Dev] I18N Toolkit
References: <19991109135839.25864.rocketmail@web607.mail.yahoo.com><199911091646.LAA21467@eric.cnri.reston.va.us><14376.22527.323888.677816@amarok.cnri.reston.va.us><199911091726.MAA21754@eric.cnri.reston.va.us><14376.23671.250752.637144@amarok.cnri.reston.va.us><14376.25768.368164.88151@anthem.cnri.reston.va.us> <14376.30652.201552.116828@amarok.cnri.reston.va.us> <027101bf2b68$6a286a90$f29b12c2@secret.pythonware.com>
Message-ID: <028501bf2b69$7a527d10$f29b12c2@secret.pythonware.com>

I wrote:
> a slightly hairer design issue is what combinations
> of pattern and string the new 're' will handle.

and posted it to the wrong list.  sorry for
the confusion.

</F>




From cfelling at iae.nl  Tue Nov  9 14:47:52 1999
From: cfelling at iae.nl (Carel Fellinger)
Date: 9 Nov 1999 20:47:52 +0100
Subject: making the Python case to my dept.
References: <9911051740130D.02530@william.jinsa.org> <7GOU3.272$Ae3.194028@news-dal.corridex.com> <801p16$10s$1@vvs.superst.iae.nl> <tlBV3.286$Ae3.202317@news-dal.corridex.com>
Message-ID: <809tp8$1ek$1@vvs.superst.iae.nl>

Evan Simpson <evan at tokenexchange.com> wrote:
> http://www.alice.org

thanks, looks good, but elas not on linux, and no room on my windows disk,
so I've to wait a few weeks

-- 
groetjes, carel



From boud at rempt.xs4all.nl  Tue Nov 30 01:59:15 1999
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: 30 Nov 1999 06:59:15 GMT
Subject: XPython (X11)
References: <383E5350.1E07536A@xrce.xerox.com> <slrn83tj53.k7h.bernhard@alpha1.csd.uwm.edu> <81ms2n$126$1@essle.valdyas.org> <slrn840dt7.mov.bernhard@alpha1.csd.uwm.edu> <m3hfi64oyh.fsf@phaedrus.zen.xom>
Message-ID: <81vsk3$g8$3@essle.valdyas.org>

Rob Hodges <s323140 at student.uq.edu.au> wrote:
> bernhard at alpha1.csd.uwm.edu (Bernhard Reiter) writes:

> I'm struggling a bit to understand why you say that in this context.
> I can't speak for pyKDE (although I imagine the situation is the
> same), but developing with pyGnome doesn't imply a 'tying' to Gnome;
> only that the Gnome libs must be available in addition to the GTK libs
> -- just as you must have wxwin in addition to GTK to run wxpython
> apps.  One advantage of PyGnome over straight PyGtk is that it offers
> the Gnome Canvas widget, which can be useful whether the user is
> running Gnome or not.

Likewise, one of the strong points of PyKDE is the KHTMLView widget,
which can be put to good use in a lot of applications. And I like
things like global key accelerators. Anyway, any app where you can't
easily substitute one gui for another has some design flaws.

-- 

Boudewijn Rempt  | http://denden.conlang.org



From jcosby at wolfenet.com  Tue Nov 23 20:32:45 1999
From: jcosby at wolfenet.com (Jon Cosby)
Date: Tue, 23 Nov 1999 17:32:45 -0800
Subject: Problems configuring Personal Web Server
Message-ID: <81ff61$c4e$1@sparky.wolfe.net>

I need help configuring MS Personal Web Server to run Python CGI scripts.
I've followed the steps given on Aaron Watters' web page
(http://starship.python.net/crew/aaron_watters/pws.html), but my server
isn't recognizing my scripts as executables; I'm being prompted to download
them instead. It worked for me in Windows 95, I'm in Windows 98 now. I can't
see any other difference. I've checked my registry editor over and over, and
everything looks right. Any guesses as to what I'm overlooking?

--
Jon Cosby

E-mail: jcosby at wolfenet.com
Website: http://www.wolfenet.com/~jcosby/





From m.faassen at vet.uu.nl  Mon Nov 15 10:38:36 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 15 Nov 1999 15:38:36 GMT
Subject: Beginning programming with Python
References: <sng0ychgwt.fsf@motorola.com> <382B3A53.5816B2@callware.com> <382B3D79.74CD3E0F@email.sps.mot.com>
Message-ID: <80p9ds$ppo$1@newshost.accu.uu.nl>

"Aaron Ginn (rp2885)" <rp2885 at email.sps.mot.com> wrote:
> An honest opinion, please; is Python better suited for beginners than Perl, as
> I've heard?

You're asking in the Python newsgroup, so here's my honest opinion:

Yes.

I was an experienced programmer (C, C++, Pascal, and some glancing 
familiarity with a host of other languages) when I learned Python.
Before I looked at Python, I struggled a few weeks trying to teach
myself Perl. Then I downloaded the Python tutorial, read it in about an
hour, and I never looked back. I could program in Python from then on. Python
was much much easier for me to learn. I could reason about the right
syntax for things, while with Perl I was constantly checking the documentation.

This was of course the example of an experienced programmer, but I think
it'll be true for beginners too. The multiplicity of syntax of Perl to
do the same thing, and the odd syntax to do certain things will confuse
beginners, and will also teach the wrong idea to beginners; programming
is not about syntax gimmicks. With perl you could be tempted to learn
lots of snippets and tricks to solve a problem and how to assemble
these to a working program. Lots of copy&paste programming, and lots of
'how do I do this in Perl' questions. With Python you can focus more on
the underlying structure, the semantics, of the program.

Then again, this is a biased opinion. There are arguments why Perl would
suit the human mind better *because* it is so syntactically variable and
complicated. I don't believe in these arguments myself.

Python has a clean and powerful syntax. Perl has just a powerful syntax.

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?



From edward at corpex.com  Wed Nov 10 03:46:42 1999
From: edward at corpex.com (Edward Hasted)
Date: Wed, 10 Nov 1999 08:46:42 GMT
Subject: Text Manipulation in Python
Message-ID: <VA.00000012.055a8702@ops.corpex.com>

I am completely new to Python.

We want to use it to alter specific lines in template files, typically 
something like changing:-

Variable = 1234

to 

Variable = Company Name

The text manipulation strings within Python seem to work sequentially 
rather than on a line basis.

1. Is this correct?
2. What is the best way to do text searching and manipulation within 
Python.

Many thanks,

Edward 




From grant.griffin at iowegian.com  Sun Nov  7 10:47:36 1999
From: grant.griffin at iowegian.com (Grant Griffin)
Date: Sun, 07 Nov 1999 15:47:36 +0000
Subject: python: something completely different
References: <3820C0B8.62B2@iowegian.com> <38249ED2.BB7EA1DE@znet.com> <8028jv$7cf$1@hub.org>
Message-ID: <38259F18.437@iowegian.com>

Patty Thomas wrote:
> 
> On Sat, 06 Nov 1999 13:34:10 -0800, Robert Cain <arcane at znet.com> wrote:
> >Back to the subject at hand (Python in the DSP universe) has anyone
> >investigated the possibility of conversion, compilation or whatever of
> >MatLab to Python?  I'm naive WRT to both at this point because 1) MatLab
> 
> I know nothing about MatLab, but a quick search here
> http://www.vex.net/~x/parnassus/apyllo.py?so=d&find=matlab reveals that it
> would seem Travis Oliphant has some MatLab like tools (or functions at least)
> which could be useful until someone 'converts, compiles, or whatever', <grin>
> matlab to python here... http://oliphant.netpedia.net/

As you suggest, Travis is creating a Matlab-like "Signal Processing
Toolbox" for Python.

I don't know if a Matlab-to-Python translator is currently available,
but if not, I don't think it would be all that hard to write one (in
Python, of course).  

=g2
-- 
_____________________________________________________________________

Grant R. Griffin                                       g2 at dspguru.com
Publisher of dspGuru                           http://www.dspguru.com
Iowegian International Corporation	      http://www.iowegian.com



From greg.ewing at compaq.com  Fri Nov 19 03:57:53 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Fri, 19 Nov 1999 21:57:53 +1300
Subject: Type/Class Unification in Ruby
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp> <7vjtrc$hmm$1@newshost.accu.uu.nl> <38204614.2EAC6051@compaq.com> <5ld7trtak5.fsf@eric.cnri.reston.va.us> <3826D63C.2731BB1F@compaq.com> <38328948.9E35556F@compaq.com> <m3d7t9v8lr.fsf@qiao.localnet>
Message-ID: <38351111.6B3E08C5@compaq.com>

Clemens Hintze wrote:
> 
> But it is worth mentioning, that Ruby features
> single inheritance *by*purpose*. Not because of the implementation
> details!

I wasn't making a value judgement, only pointing out that
Ruby's solution is as simple as it is because of this
restriction. 

> If it would implement multi-inheritance, the class structure could
> e.g. use an array of pointer to all such C structs, not only one
> pointer, as implemented now!

Sure, Python could do something like that too. But it would
get very hairy and complicated and probably slow everything
down a lot, casting doubt on whether it's worth the effort.

A middle ground might be to allow inheriting from any number
of user-defined classes, but at most one built-in type.

> The second (preferred) way is, to build instances of class DATA
> wrapping the C struct you have allocated for your extension.

Yes, I know, but what I was saying still applies in that case.
All the methods of the extension class expect the DATA object
to be pointing to their own particular kind of struct, and
because of the inheritance structure, this is always true --
so they don't need to bother checking.

> Deriving a Ruby class coded in C from another one also coded in C,
> seems to be the only right tricky part of that language, IMHO. :-)

I think it will work as long as the derived class makes sure
that its C structure is an exension of the superclass's C
structure.

>     Greg> [global instance var dict] might cause ref counting
>     Greg> problems.
> 
> Once again, thanks for Ruby's true garbage collection! :-)

Garbage collection alone doesn't actually solve the problem;
the references from the global dict need to be treated as
weak references somehow under either scheme. I haven't
looked, but presumably Ruby must be ignoring the global dict
refs when marking, and cleaning them up whenever a builtin
or extension object is deallocated. Python could be made to
do the equivalent.

> The big exception,
> AFAIK, was class Fixnum. It never had any instanciation method, as it
> doesn't wrap any struct.

Yes, that's the one class you really can't instantiate a
subclass of, since its class is totally implicit.

But it's not clear whether it's any use having something
like a UserFlonum, either, even though technically it could
be done. Unless you override every method, any arithmetic
on it would return an ordinary Flonum. And if you are going
to override every method, it might as well be a totally
new class!

> Often it is better to use the module
> mechanism to add methods to class Array or some of its instances
> during runtime.

You mean you can add methods to an *existing* class that
way? Not sure if I like that idea...

Greg



From endgamer71 at hotmail.com  Tue Nov  9 08:52:46 1999
From: endgamer71 at hotmail.com (Endgamer)
Date: Tue, 09 Nov 1999 13:52:46 GMT
Subject: Another bloody newbie question: Executable Python files?
Message-ID: <382821f1.96924220@212.225.36.99>

Is it possible (I know it's not required) to compile a Python program
to be stand-alone? Or is it required that Python be installed on every
machine the program runs on? I'm working with Windows machines, and
I'm writing this little script I need to have run on each machine (of
about 20) at startup - I'd prefer not to have to install Python on
each machine!

Thanks in advance...


Endgamer
Spider Warrior

(Endgamer71 at spamtrap.hotmail.com)
#####################################
Not a quote, just an on-the-spot observation:
You know you're a geek when your PC costs exactly
15 times as much as your stereo. And gives better
sound quality.<sigh>
                              Endgamer
#####################################



From a.eyre at optichrome.com  Wed Nov 24 12:25:23 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Wed, 24 Nov 1999 17:25:23 -0000
Subject: SMTP?
In-Reply-To: <19991124171120.1623.qmail@hotmail.com>
Message-ID: <003301bf36a0$e3bf3710$3acbd9c2@peridot.optichrome.com>

> send: 'mail FROM:<alexis> size=10\015\012'
> reply: '250 <alexis>, Sender accepted\015\012'
> reply: retcode (250); Msg: <alexis>, Sender accepted
> send: 'rcpt TO:<ulf.engstroem at nebb.no>\015\012'
> reply: '250 <ulf.engstroem at nebb.no>, Recipient ok\015\012'
> reply: retcode (250); Msg: <ulf.engstroem at nebb.no>, Recipient ok
> send: 'data \015\012'
> reply: '354 Enter mail, end with <CRLF>.<CRLF>\015\012'
> reply: retcode (354); Msg: Enter mail, end with <CRLF>.<CRLF>
> data: (354, 'Enter mail, end with <CRLF>.<CRLF>')

In here you might need to put in standard RFC822 headers.
From: xxx at yyy
To: aaa at bbb
Subject: Not a lot
[Blank line here to indicate end of headers]
Message here

> send: 'Small msg\015\012'
> send: '\015\012.\015\012'
> reply: '250 Ok, message saved\015\012'
> reply: retcode (250); Msg: Ok, message saved
> data: (250, 'Ok, message saved')
> send: 'quit \015\012'
> reply: '221 See ya in cyberspace\015\012'
> reply: retcode (221); Msg: See ya in cyberspace

Check the RFC, but I think the only important bit is the blank
line to indicate end of headers.

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From viruswall at bam.com  Thu Nov 25 16:24:54 1999
From: viruswall at bam.com (viruswall at bam.com)
Date: Thu, 25 Nov 1999 16:24:54 -0500 (Eastern Standard Time)
Subject: InterScan NT Alert
Message-ID: <199911252130.QAA13064@python.org>

Sender, Bell Atlantic Mobile has detected virus(es) in your e-mail attachment. Please take the appropiate action to protect your workstation with virus protection software or install the latest virus signature file.

Date:  	Thu, 25 Nov 1999 16:24:54 -0500 (Eastern Standard Time)
Method:	Mail
From:  	<python-list at python.org>
To:    	<python-list at python.org>
File:  	Happy99.exe
Action:	clean failed - deleted
Virus: 	TROJ_SKA 



From mikael at isy.liu.se  Fri Nov  5 04:31:03 1999
From: mikael at isy.liu.se (Mikael Olofsson)
Date: Fri, 05 Nov 1999 11:31:03 +0200 (MET)
Subject: __str__ vs. __repr__
In-Reply-To: <19991105102233.I29354@xs4all.nl>
Message-ID: <XFMail.991105113103.mikael@isy.liu.se>

On 05-Nov-99 Thomas Wouters wrote:
 >  We're too busy bitching about the idiocy of people who expect to change
 >  a country's (usually) ancient currency, habits and not to mention every
 >  keyboard, font and accountancy software in the world in a matter of years.
 >  Or perhaps we just dont _care_ about the euro symbol. Your Euro symbol, by
 >  the way, Tim, ended up as '?' in my font -- but possibly because isprint()
 >  returned 0 and my mail-reader therefor refused to print it ;)

Actually, many European currecies are not more than approximately 100
years old, and so are many of the countries as well. That's not so
ancient to me. But it's true; many of us don't care.

/Mikael

-----------------------------------------------------------------------
E-Mail:  Mikael Olofsson <mikael at isy.liu.se>
WWW:     http://www.dtr.isy.liu.se/dtr/staff/mikael
Phone:   +46 - (0)13 - 28 1343
Telefax: +46 - (0)13 - 28 1339
Date:    05-Nov-99
Time:    11:26:19

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



From cfelling at iae.nl  Mon Nov  8 17:06:12 1999
From: cfelling at iae.nl (Carel Fellinger)
Date: 8 Nov 1999 23:06:12 +0100
Subject: a simple helper class for parsing command line options
References: <s1sj8n.11uoa9u@dev.huh.harvard.edu>
Message-ID: <807hgk$289$1@vvs.superst.iae.nl>

In article <s1sj8n.11uoa9u at dev.huh.harvard.edu> you wrote:
> I threw this together to make parsing command line options and arguments
> a little easier than just using getopt directly.  Feedback welcome.

Sorry for reacting so late, but it took some time to cough up a responce.
(or actually to build my own option state machine)
Besides being publically accused of Perlish behaviour
made me hesitate to react:), but now that no one else has...

A bunch of nice ideas: inspired me to adjust it, rewrite it and
eventually write my own:), But what would you expect with something
so close to the look of ones program and moreover from a Dutch man.

> class OptionFactory:

this triggered me to see option parsing as a state machine

>         self.__options = [('help','h','help','','print this help message')]

I personally have some problems reading all those quotes, so I would
prefer something more friendly like:

    self.__options = [('-h, --help : print this help message')]

>         self.__dict = {}

a seperate dict for the state, I keep that:)
(though I would call it __state, less likely to be confused with __dict__)

>     def __getitem__(self, item):
>         return self.__dict[item]

and these, ofcourse

>     def add_options(self, options):
>         """
>         label = identifier for the option, used to retrieve option value

my problem here is, why should I specify an extra identifier, I could as
well use the short and or long flags themselves and enhance __getitem__
to deal with that.

>         value = descriptive string of the value the option requires

it took a while before I stopt forgetting that you refer here to the
optional argument to the option flags and not to the actual value
specified on the command line. Well that just shows how silly I am.

>     def add_argdesc(self, argdesc):
>         """ add a description of the expected arguments """

Great. I would add a similar one to allow for the replacement of the
default [OPTIONS...] string as well as it is hard to construct a
meaningfull option list automagically.

>     def usage(self):

a lot of code, necessary to join the flags part of the description.
In my previously proposed variant, this flags part could be used almost
without any massaging. And what about multiline descriptions?

>     def parse(self, argv):
> 	""" pass in sys.argv, not sys.arv[1:] """

or define a default value like:

    def parse(self, argv=argv[1:]):

>         self.name = argv[0]

definitatively

>         try:
>             optlist, args = getopt(argv[1:], self.__shortstring, self.__longlist)
>             for opt in optlist:
>                 self.__process_opt(opt)
>             self.args = args
>         except error, x:
>             msg = 'Error parsing options:  %s' % x
>             sys.stderr.write('%s\n' % msg)
>             sys.stderr.write('\n%s\n' % self.usage())
>             sys.exit(1)

nice to catch getopt.error here, though I myself would prefer a short
responce to stderr refering to the use of the --help option. Just hate
to have to scroll back just to see what I was doing:)


For those interested: I redid the above, treating option parsing more
like transitions in a state machine. Using that module one can say
things like:

    options = OptionState(test=0, verbosity=0)
    add = options.add_option
    add('-t, --test : just testing', test=1)
    add('--version : show the version and exit', options._version)
    add('--help : show this help message', options._help)
    add('-v : increment the verbosity', options._inc, 'verbosity')
    add('-V : increase verbosity with 5', options._inc, 'verbosity', 5)
    add('-D : set verbosity to 11', options._int, 'verbosity', 11)
    add('-x : ', x=1) # secret option, not displayed in help message!
    ...
    if options.verbosity > 8: pass
    if options['test']: pass
    if options.x or options.has_option('x'): pass
    
One can also keep a list of values of repeated option flags, or easily
enhance this state machine with your own special purpose transitions.

-- 
groetjes, carel
-- 
groetjes, carel



From mal at lemburg.com  Mon Nov  1 04:56:26 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Mon, 01 Nov 1999 10:56:26 +0100
Subject: os.popen() and (IDLE, PythonWin) ?
References: <NDBBIAJJDMLKMBCHENJJMEAACEAA.kputland@wisen.com> <3819DD05.FF6656C3@lemburg.com>
Message-ID: <381D63CA.9073803D@lemburg.com>

"M.-A. Lemburg" wrote:
> 
> [os.popen('ver') fails under IDLE and PythonWin, yet works under "normal"
>  Python]

Ok, I found the FAQ entry on this:

7.13. Why doesn't os.popen() work in PythonWin on NT?

The reason that os.popen() doesn't work from within PythonWin is due to a bug in Microsoft's C Runtime
Library (CRT). The CRT assumes you have a Win32 console attached to the process. 

You should use the win32pipe module's popen() instead which doesn't depend on having an attached Win32
console. 

Example: 

 import win32pipe
 f = win32pipe.popen('dir /c c:\\')
 print f.readlines()
 f.close()

Under PythonWin this is no problem (although I wish os.popen would
automatically choose the different approach for me in the above
setup). Yet under IDLE I have a problem: win32 doesn't have to be
installed... is there some other workaround ?

-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                    60 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/



From mwh21 at cam.ac.uk  Tue Nov 30 10:31:41 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 30 Nov 1999 15:31:41 +0000
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <81bmmi$279$1@nnrp1.deja.com> <3841EE77.6B5D6AC9@wjk.mv.com> <FLxz96.IEI@world.std.com> <38435C43.B238EC15@mindspring.com> <FM0KtD.2ps@world.std.com>
Message-ID: <m33dtoro6a.fsf@atrus.jesus.cam.ac.uk>

wware-nospam at world.std.com (Will Ware) writes:

> Will Ware wrote:
> > I find myself grumbling about having to type "x = x + 1". The really
> > clean thing to do, given that integers are objects, would be to define
> > increment and decrement methods, so you'd type something like "i.incr()".
> 
> Jesse D. Sightler (jsight at mindspring.com) wrote:
> > Wouldn't this be impossible, since Integers are also immutable?  So, eg,
> > i.incr() could only return an incremented version of i, but not actually
> > increment i?
> 
> Right, this is what Gordon was pointing out. A "++" method could not
> magically reach out of its object, find the "i" variable, and bind it
> to the incremented value. 

Hate to be awkward, but when it comes to things like this, there's
very little like this that can't be done in Python!

>>> def incr(var):
	try:
		raise 1/0
	except ZeroDivisionError:
		import sys,dis
		f = sys.exc_traceback.tb_frame.f_back
		loadins = f.f_code.co_code[f.f_lasti - 3:f.f_lasti]
		loadop = ord(loadins[0])
		name = dis.opname[loadop]
		loadarg = ord(loadins[1]) + 256*ord(loadins[2])
		vname = f.f_code.co_names[loadarg]
		f.f_locals[vname] = f.f_locals[vname] + 1

		
>>> x=1
>>> incr(x)
>>> x
2

well, it works in the IDLE shell window...

I'll stop now ;-).

> Something like this could be accomplished by tinkering with Python's
> parser, but that wouldn't be a great idea.

As above, not necessary.

> A lot of work for something that would never find wide acceptance.

That much is true!

Cheers,
Michael



From usenet-mm-yyyy at julesallen.com  Wed Nov 10 19:51:43 1999
From: usenet-mm-yyyy at julesallen.com (Jules Allen)
Date: Thu, 11 Nov 1999 00:51:43 GMT
Subject: IPC8: Call the 800 number!
References: <3826f6ac.0@news.bestsoftware.com> <SjnW3.306$0O.5636532@news.randori.com>
Message-ID: <zqoW3.1431$AG3.24638@dfiatx1-snr1.gtei.net>

In article <SjnW3.306$0O.5636532 at news.randori.com>, Robin Dunn wrote:
| We called the 800 number and they still didn't know about it.  They did
| however have an $84 room that I snached up faster than they could say "Would
| you like that one?"  Now I just hope it's in the right hotel and not next to
| the dumpster or something...

Actually that $84 room *is* a dumpster...

-- 
Want to mail me? Put the current numeric month and year in the 
obvious place.



From Gareth.McCaughan at pobox.com  Thu Nov 25 15:52:26 1999
From: Gareth.McCaughan at pobox.com (Gareth McCaughan)
Date: 25 Nov 1999 20:52:26 +0000
Subject: Python complaints
References: <1268673769-3409210@hypernet.com> <86puwz4jhp.fsf@g.local>
	<14397.6317.426060.330244@gargle.gargle.HOWL>
Message-ID: <86iu2q2t11.fsf@g.local>

Charles Waldman wrote:

> You can often (ab)use dictionaries to get a ?:-like effect, as in this 
> construction:
> 
> def p(score):
>     print "You scored %d point%s" % (score, {1:''}.get(score,'s'))

That's disgusting. But very neat. :-)

-- 
Gareth McCaughan  Gareth.McCaughan at pobox.com
sig under construction



From nbecker at fred.net  Mon Nov  1 13:20:01 1999
From: nbecker at fred.net (nbecker at fred.net)
Date: 01 Nov 1999 13:20:01 -0500
Subject: Need Berkley DB!!
References: <381D5E84.98931682@bibsyst.no> <14365.51195.136678.559903@dolphin.mojam.com>
Message-ID: <x88yaci846m.fsf@rpppc2.hns.com>

You shouldn't need to compile db - your linux already comes with it.

>From /usr/include/db.h:
#define	DB_VERSION_STRING	"Sleepycat Software: DB 2.4.14: (6/2/98)"



From MHammond at skippinet.com.au  Fri Nov  5 17:29:27 1999
From: MHammond at skippinet.com.au (Mark Hammond)
Date: Fri, 05 Nov 1999 22:29:27 GMT
Subject: Accessing Word object model from Python COM server
References: <3822a7d2.251040356@news.mch.sni.de>
Message-ID: <bTIU3.1698$pf4.10851@news-server.bigpond.net.au>

Nikolai Kirsebom wrote in message <3822a7d2.251040356 at news.mch.sni.de>...
>
>When the GetNew method in the Python COM server is executed, an error
>indicates that the name 'Documents' is invalid.  The self.ApplObj is
>of type PyIDispatch (I think), however may it must be an 'instance' in
>order to be able to access the object model.  If I instead of

This is correct.  What you need to do is

doc = win32com.client.Dispatch(doc)

This will wrap it in the Python class that manages all the magic.

Mark.






From fredrik at pythonware.com  Sat Nov 20 06:03:37 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 20 Nov 1999 12:03:37 +0100
Subject: Displaying an array as an image
References: <m33du2m4w4.fsf@c34680-a.grlnd1.tx.home.com>
Message-ID: <006101bf3346$e6264be0$f29b12c2@secret.pythonware.com>

David Fenyes <dfenyes at flash.net> wrote:
> Is there an easy way to display a 2D array as an image using just
> Tkinter?  I'm aware of a TK Image canvas widget, but can't find docs
> on how to use it or what an 'image' is supposed to consist of.  I'm
> just evaluating numerical python as a platform and this is a pretty
> important functionality that is not really addressed in any of the
> readily available docs.

see the attached message.  and remember that
python.org's search facility is your friend.

> PIL and friends seem to be more oriented to reading/writing image
> formats, and are not really what I'm looking for.

since you obviously don't know what PIL can do for
you, how can you be so sure?

check out:

http://www.pythonware.com/library/pil/handbook/imagetk.htm

and the demo scripts in the source distribution.

</F>

Newsgroups: comp.lang.python
From: "Fredrik Lundh" <fredrik at pythonware.com>
Subject: Re: Tkinter PhotoImage Question
Date: Fri, 12 Nov 1999 16:06:38 +0100

Ivan Van Laningham <ivanlan at callware.com> wrote:
> Ok, I've managed to create an image, using xx=PhotoImage(file), and I
> can retrieve pixel values from that image using xx.get(x,y).
> 
> Now I want to use the put() method to write a pixel.
> 
> How do I do that?  I would have expected that put() would take a string
> of pixel values for the data argument, but that doesn't seem to be the
> case. ...

from the eff-bot archives:

#
# create a greyscale ramp using pure Tkinter
#
# fredrik lundh, january 1998
#
# fredrik at pythonware.com
# http://www.pythonware.com
#

from Tkinter import *

# must initialize interpreter before you can create an image
root = Tk()

data = range(256) # 0..255

im = PhotoImage(width=len(data), height=1)

# tkinter wants a list of pixel lists, where each item is
# a tk colour specification (e.g. "#120432").  map the
# data to a list of strings, and convert the list to the
# appropriate tuple structure

im.put( (tuple(map(lambda v: "#%02x%02x%02x" % (v, v, v), data)),) )

# resize the image to the appropriate height

im = im.zoom(1, 32)

# and display it

w = Label(root, image=im, bd=0)
w.pack()

mainloop()

# after playing with this a little, you'll love
# the Python Imaging Library.




From skip at mojam.com  Wed Nov 24 14:10:25 1999
From: skip at mojam.com (Skip Montanaro)
Date: Wed, 24 Nov 1999 13:10:25 -0600 (CST)
Subject: whitespace.py module
In-Reply-To: <y7b66yrdel7.fsf@agastya.eng.fore.com>
References: <y7b66yrdel7.fsf@agastya.eng.fore.com>
Message-ID: <14396.14369.586465.200108@dolphin.mojam.com>

    Rajesh> I'm new to the python language, and the newsgroup. So, if this
    Rajesh> newsgroup is not appropriate for this type of posting, please
    Rajesh> pardon my ignorance and notify me.

Perfectly appropriate.

    Rajesh> I have written up a module called `whitespace.py' available at
    Rajesh> <URL:http://www.dsmit.com/python/>

    Rajesh> and a sample program `whitespace' which uses this module, also available at
    Rajesh> the same URL.

Have you checked out tabnanny in Tools/scripts/tabnanny.py of the Python
distribution?  A merging of interests might be useful.  I suspect the two
scripts detect similar problems.

Skip Montanaro | http://www.mojam.com/
skip at mojam.com | http://www.musi-cal.com/
847-971-7098   | Python: Programming the way Guido indented...




From mcfletch at vrtelecom.com  Wed Nov 17 00:06:45 1999
From: mcfletch at vrtelecom.com (Mike Fletcher)
Date: Wed, 17 Nov 1999 00:06:45 -0500
Subject: list.without()?
In-Reply-To: <19991116221403.D7366@krikkit.vollmert.home>
Message-ID: <NDBBIAPHEKGOBIHCBLIICEPBCAAA.mcfletch@vrtelecom.com>

Nope, not a big difference.  Requires a few extra steps if the original list
is [], testing for list "truth" is pretty fast.

Cheers,
Mike

-----Original Message-----
From: python-list-admin at python.org
[mailto:python-list-admin at python.org]On Behalf Of Robert Vollmert
Sent: November 16, 1999 4:14 PM
To: python-list at python.org
Subject: Re: list.without()?

...

Would it make a big difference if
  while temp:
were replaced by
  while 1:
?
...




From shouldbe at message.com  Tue Nov 16 14:24:18 1999
From: shouldbe at message.com (Arinte)
Date: Tue, 16 Nov 1999 14:24:18 -0500
Subject: Final classes like in java???
Message-ID: <80sao7$1848$1@rtpnews.raleigh.ibm.com>

Is there a way you can define your python class to be final or not able to
be inherited from?





From fredrik at pythonware.com  Tue Nov 30 03:45:04 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 30 Nov 1999 09:45:04 +0100
Subject: Embedded size?
References: <3842cf05.407909516@news.earthlink.net>
Message-ID: <022501bf3b14$9445a480$f29b12c2@secret.pythonware.com>

freshhorses at earthlink.net wrote:
> Were I to embed Python as my in-application scripting language, how
> much space would it need in both code and memory?  I don't see
> anything obvious on the org site discussing those sorts of
> requirements :)

well, if you do things the right way, you'll
probably save both code and memory by
adding the interpreter to your app (not to
mention development time!).

but if you're interested in downsizing python,
take a look at the material at:

    http://www.abo.fi/~iporres/python/

</F>




From smigowsky at dspace.de  Fri Nov 19 06:58:16 1999
From: smigowsky at dspace.de (Stefan Migowsky)
Date: Fri, 19 Nov 1999 12:58:16 +0100
Subject: Running Exel
Message-ID: <326536345498D311B3BC00105A39802A0744E8@PDC-DSPACE>

See file :

...\Python\win32com\test\testMSOffice.py 

There are examples how to use the office suite.

Stefan


>-----Original Message-----
>From: Przemyslaw G. Gawronski [mailto:gawron at obop.com.pl]
>Sent: Friday, November 19, 1999 12:16 PM
>To: python-list at python.org
>Subject: Running Exel
>
>
>How can I run MS Exel via OLE using Python (with windows ext.) ?
>
>Thankx
>
>Przemek
>
>--
>Przemyslaw G. Gawronski UIN:8358522
>mailto:gawronskip at usa.net
>mailto:gawron at obop.com.pl
>
>
>
>-- 
>http://www.python.org/mailman/listinfo/python-list
>



From fredrik at pythonware.com  Mon Nov 22 10:21:04 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 22 Nov 1999 16:21:04 +0100
Subject: CRC-module
References: <38396766.B189D15A@bibsyst.no>
Message-ID: <006c01bf34fd$31ce5800$f29b12c2@secret.pythonware.com>

Thomas Weholt <thomas at bibsyst.no> wrote:
> Is there a crc-module or an easy way to compute crc-values from files? I
> want to store these values in a database to check ftp-sites, cd-roms
> etc. to see if there are updated version or new files etc. ?

here's a few:

    zlib.crc32 (32-bit checksum)
    zlib.adler32 (32-bit checksum)
    md5 (128-bit digest)
    sha (160-bit digest)

see the library reference (or the eff-bot
guide) for more info.

</F>

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




From fdrake at acm.org  Tue Nov 23 10:52:43 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Tue, 23 Nov 1999 10:52:43 -0500 (EST)
Subject: Python books
In-Reply-To: <383A9EF4.5D1E8986@callware.com>
References: <199911230613.HAA21179@dionysus.fw.cuci.nl>
	<383A9EF4.5D1E8986@callware.com>
Message-ID: <14394.47179.621762.615067@weyr.cnri.reston.va.us>

Ivan Van Laningham writes:
 > If "Python Annotated Archives" is /F's book, it's already out as an
 > eMatter book.  Visit www.python.org for details on how to get it. 
 > Search Deja News for important information on using this style of
 > book--there was a thread on here about it.

  It's not -- this is something from McGraw-Hill.  I don't think it's
out yet.


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From tseaver at starbase.neosoft.com  Mon Nov 15 15:48:39 1999
From: tseaver at starbase.neosoft.com (Tres Seaver)
Date: 15 Nov 1999 20:48:39 GMT
Subject: __str__ vs. __repr__
References: <000001bf275c$02370d20$1d2d153f@tim> <01fd01bf2dce$d2c03cf0$f29b12c2@secret.pythonware.com> <942499879snz@vision25.demon.co.uk> <m3n1shp2uu.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <B65809622FDC7452.143B25D04FC64B07.A4A6F24F2451314F@lp.airnews.net>

In article <m3n1shp2uu.fsf at atrus.jesus.cam.ac.uk>,
Michael Hudson  <mwh21 at cam.ac.uk> wrote:
>philh at vision25.demon.co.uk (Phil Hunt) writes:
>
>> In article <01fd01bf2dce$d2c03cf0$f29b12c2 at secret.pythonware.com>
>>            fredrik at pythonware.com "Fredrik Lundh" writes:
>> > 2) what's the reason byte code cannot be used to recover
>> > the original source code
>> 
>> Perhaps because two different (but obviously functionally
>> identical and very similar) source codes could produce the same
>> byte code.
>
>I doubt it actually. Python's bytecode is really pretty high
>level. Leastaways, I can't think of any examples.
>
>It's one of those things I might get round to writing one of these
>days. I don't really know how hard it would be. Not especially, I
>suspect.

Besides, you need to defer it until _after_ this year's Pythonic award,
for which you are already in pole position for inventing bytecodehacks :).

Tres.
-- 
---------------------------------------------------------------
Tres Seaver           tseaver at palladion.com       713-523-6582
Palladion Software    http://www.palladion.com



From a.eyre at optichrome.com  Wed Nov 24 04:52:36 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Wed, 24 Nov 1999 09:52:36 -0000
Subject: Multi-Dimensional Arrays.
Message-ID: <002c01bf3661$a2e27160$3acbd9c2@peridot.optichrome.com>

> Allow me to refine the question. If I make a list of lists to be my
> multi-dimensional array, how do I pass that into C?
> PyArg_ParseTuble(args, "?", &x)?

That should work.

> What would go in place of the "?"?

A big 'O'.

> Or would it be done a different way.  Then to pass it back to Python,
> I would return what?  Could I use Py_BuildVal("?", x) ? 

You can.

The PyArg_Parse and Py_BuildValue are really for converting to/from
C types. Since there is no obvious conversion for a Python list,
you must use the Python list API (function beginnig with PyList.)

Something like:

PyObject* myfunc(PyObject* self, PyObject* args)
{
	PyObject* mylist;
	int length, c;
	
	if (!PyArg_ParseTuple(args, "O", &mylist))
		return NULL;

	length = PyList_Size(mylist);

	for (c = 0; c < length; c++)
	{
		PyObject* item = PyList_GetItem(mylist, c);
		/* Do something with item */
	}

	/* Let's return the empty list */
	return PyList_New(0);
}

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From mhammond at skippinet.com.au  Mon Nov 29 23:56:31 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Tue, 30 Nov 1999 04:56:31 GMT
Subject: Pythonwin Reset
References: <01bf38df$8127e0c0$3a42c3d1@Cyberus.cyberus.ca> <_X004.7432$VA6.38559@news-server.bigpond.net.au> <01bf39bd$a28831a0$4155c3d1@Cyberus.cyberus.ca> <81rmh5$9e$1@hub.org> <01bf3a8c$3f855b80$2241c3d1@Cyberus.cyberus.ca>
Message-ID: <3OI04.9647$VA6.47885@news-server.bigpond.net.au>

Gordon Williams wrote in message
<01bf3a8c$3f855b80$2241c3d1 at Cyberus.cyberus.ca>...
>Tim,
>
>Thanks for the note.
>
>It now works in that it eliminates all the names, but it doesnt quite fix
>my problem.  When I make a GUI with wxPython (using Pythonwin IDE) and I
>get an exception, the next time I try and run the program after fixing the
>mistake, I get another exception saying One wxApps process only!  Something
>was left behind.  The only way I have found to clear the problem was to
>close Pythonwin and then restart it and reload the file.


This is a known issue with Pythonwin and wxPython.  I believe the issue is
mainly in wxPython and is on the (no doubt) long list of things to be
done...

Mark.





From fdrake at acm.org  Wed Nov 17 09:49:05 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Wed, 17 Nov 1999 09:49:05 -0500 (EST)
Subject: docs
In-Reply-To: <000501bf30d9$a6e93100$a42d153f@tim>
References: <14385.28690.800102.134431@weyr.cnri.reston.va.us>
	<000501bf30d9$a6e93100$a42d153f@tim>
Message-ID: <14386.49249.39479.67579@weyr.cnri.reston.va.us>

Tim Peters writes:
 > No, I doubt the service pack has anything to do with it.  What's "obvious"
 > in the MS world is that it's a random bug that maybe 1 person in 1000 can
 > reproduce.  That's par for the MS course!  Or did many people report the
 > same problem with NT SP4?  I saw one report here, and a partial possible

  Only the one report, but that's enough for me to implement a simple
work-around.
  As a Unix-weenie, I consider it evil to reformat & reinstall an
entire system!  Heck, I don't even restart XEmacs for months if the
power company cooperates.  ;-)

 > The usual MS fix is to scandisk, defrag, clear all the caches, and reinstall
 > stuff from scratch.  And I haven't had more than 8 browser bugs that weren't
 > eventually cured by reformatting my hard drives <wink>.

  I fixed an NT installation by repartitioning and reformatting once,
but I skipped the reinstallation step and used the space under my O/S
of choice.  Works better now.  ;-)


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From cfelling at iae.nl  Tue Nov  9 14:45:47 1999
From: cfelling at iae.nl (Carel Fellinger)
Date: 9 Nov 1999 20:45:47 +0100
Subject: Python for children.....
References: <9911051740130D.02530@william.jinsa.org> <7GOU3.272$Ae3.194028@news-dal.corridex.com> <801p16$10s$1@vvs.superst.iae.nl> <199911080328.WAA24200@acer.spvi.com> <199911080333.WAA24231@acer.spvi.com>
Message-ID: <809tlb$1eg$1@vvs.superst.iae.nl>

Steve Spicklemire <steve at acer.spvi.com> wrote:

> Forgot to mention:

> PyBots: http://opensource.spvi.com

looks interesting, but elas, it doesn't run on linux
and i've hardly any room on my left-over windows partition.
so I've to wait for an other port or Sinterklaas.

-- 
groetjes, carel



From kuncej at mail.conservation.state.mo.us  Wed Nov 10 18:50:18 1999
From: kuncej at mail.conservation.state.mo.us (Jeffrey Kunce)
Date: Wed, 10 Nov 1999 17:50:18 -0600
Subject: Shared NT network installation of python
Message-ID: <s829b088.089@mail.conservation.state.mo.us>

I've worked out a procedure to run python (including tkinter) on NT 
from a shared network installation.

It is working for me with only some minor glitches. I'm making it available 
with the ***CAUTION*** that it is an experimental work in progress. 
If it trashes your registry, don't blame me!

See:
    pynetinst
at:
    http://starship.python.net/crew/jjkunce/ 

Comments and suggestions are welcome.

  --Jeff




From herzog at online.de  Thu Nov  4 10:13:12 1999
From: herzog at online.de (Bernhard Herzog)
Date: 04 Nov 1999 16:13:12 +0100
Subject: __str__ vs. __repr__
References: <000401bf25b9$88ea57e0$2e2d153f@tim> <199911040157.UAA11908@eric.cnri.reston.va.us> <m31za6339e.fsf@greebo.nodomain.de> <00c001bf26c9$14f75a20$4500a8c0@thomasnotebook>
Message-ID: <m3d7tqqohj.fsf@greebo.nodomain.de>

"Thomas Heller" <thomas.heller at ion-tof.com> writes:

> > How about a hook function a la __import__? I.e. have a function
> > __format__ [1] in __builtin__ that is bound to repr by default. Whenever
> > Python prints the result of an expression in interactive mode, it calls
> > __format__ with the result as parameter and expects it to return a
> > string ready for printing.
> 
> But you have a custom python interpreter, don't you?

No, I haven't.

> Or how do you
> manage to use your __format__ as the printer for interactive mode?

My phrasing might have been a bit misleading. I meant it as a suggestion
how to deal with the different user preferences when it comes to format
the result in interactive mode for printing. The python interpreter
would have to be changed if my suggestion were adopted.


-- 
Bernhard Herzog	  | Sketch, a drawing program for Unix
herzog at online.de  | http://www.online.de/home/sketch/



From aahz at netcom.com  Mon Nov  1 08:38:14 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 1 Nov 1999 13:38:14 GMT
Subject: Comparison between Python and "Ruby"
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp> <7vjtrc$hmm$1@newshost.accu.uu.nl>
Message-ID: <7vk546$g08$1@nntp1.atl.mindspring.net>

In article <7vjtrc$hmm$1 at newshost.accu.uu.nl>,
Martijn Faassen <m.faassen at vet.uu.nl> wrote:
>
>What-does-TMTOWTDI-stand-for?-ly yours,

It's the Perl motto: "There's More Than One Way To Do It"
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From cove at shell10.ba.best.com  Sat Nov 27 19:17:26 1999
From: cove at shell10.ba.best.com (Cove Schneider)
Date: 28 Nov 1999 00:17:26 GMT
Subject: fcntl/ioctl and O_NONBLOCK
References: <slrn840205.sn4.hildeb@stahlw06.stahl.bau.tu-bs.de> <slrn83t5mv.bhk.hildeb@stahlw06.stahl.bau.tu-bs.de> <81ml5t$eer$0@216.39.151.169> <81p32a$m14$1@216.39.151.169> <slrn8405je.q0.hildeb@stahlw06.stahl.bau.tu-bs.de>
Message-ID: <38407496$0$222@nntp1.ba.best.com>

Ralf Hildebrandt <hildeb at www.stahl.bau.tu-bs.de> wrote:
[snip]
> This is also fixed now, but now the popen3 Object generates about 1 MB
> of output (which I read using select() and read()), and then it STILL
> doesn't return. No crash, no exit. And once more, in top it says
> "sleep" once more. 
> Any ideas what's going on there?

Humm.. Sounds like your not checking for an EOF..

Cove

(at nbn dot com)
-- 
Cove Schneider



From fdrake at acm.org  Fri Nov 12 13:56:21 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Fri, 12 Nov 1999 13:56:21 -0500 (EST)
Subject: docs
In-Reply-To: <80hkju$okf@mail.psy.uva.nl>
References: <80epm7$vbo$1@nnrp1.deja.com>
	<14378.61648.633649.431232@weyr.cnri.reston.va.us>
	<80hkju$okf@mail.psy.uva.nl>
Message-ID: <14380.25301.616025.322479@weyr.cnri.reston.va.us>

Ionel Simionescu writes:
 > I have seen it before and I have just checked it again.
 > 
 > Just open a doc,
 > say C:\Python\Doc\tut\tut.html,
 > with Internet Explorer 5.00.2314.1003,
 > on Windows NT 4.0 SP4,
 > and refresh it (F5).

  I don't see this when using NT4 SP3; can anyone check this for other 
service packs?
  I'm really not sure how to approach this problem.  Suggestions from
people more NT-aware than myself?


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From aahz at netcom.com  Tue Nov 16 11:55:55 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 16 Nov 1999 16:55:55 GMT
Subject: Python module to format text into columns?
References: <80oq13$ql3$1@nnrp1.deja.com> <y0jogcvic4t.fsf@vier.idi.ntnu.no> <80rcj3$n23$1@nnrp1.deja.com>
Message-ID: <80s2ar$nnn$1@nntp9.atl.mindspring.net>

In article <80rcj3$n23$1 at nnrp1.deja.com>,  <hamish_lawson at yahoo.co.uk> wrote:
>
>but I'm inclined to provide (at least initially) just the direct
>methods:
>
>   formatter.setcolumn(column=1, margin=4, width=16, align=left)
>   formatter.setcolumn(column=2, margin=2, width=4, align=left)
>   formatter.setcolumn(column=3, margin=2, width=6, align=right)

Instead of explicitly naming the columns, I'd suggest passing in a list
of dicts (dicts so that you can set global defaults).  That means that
for three columns all taking in the global defaults, you'd pass in

   formatter.setcolumns ( [ {}, {}, {} ] )
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From bjorn at roguewave.com  Mon Nov  1 13:59:14 1999
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Mon, 1 Nov 1999 11:59:14 -0700 
Subject: dynamically generating function assignent
Message-ID: <51C9F3C392D0D111A15600A0C99E64A9AF5605@localhost>

You can do:

	fn = eval('func_' + 'a') # look it up in the environment
	fn()

-- bjorn

> -----Original Message-----
> From: anantk at my-deja.com [mailto:anantk at my-deja.com]
> Sent: Monday, November 01, 1999 11:19 AM
> To: python-list at python.org
> Subject: dynamically generating function assignent
> 
> 
> Hi,
> 
>   Here is my problem. I have defined say 12 pairs of functions
> 
> func_a()
> parse_func_a()
> 
> func_b()
> parse_func_b()
>  .
> .
> func_l()
> parse_func_l()
> 
>   A file contains a series of strings which are the names of these
> functions eg:
> 
> a
> c
> f
> b
> 
>   Now for each of these strings I need to call the two functions
> func_<x>(), and parse_func_<x>(), where <x> is the name of 
> the function
> that string refers to.
> 
>   I know you can do something like
>   call = func_a, and call effectively gets assigned to func_a. My
> question is how can I generate the RHS of this assignment dynamically?
> So that all I have to do is
> 
>   <variable> = <next line in file>
>   call1 = func_<variable>
>   call2 = parse_func_<variable>
>   call1()
>   call2()
> 
>   Any help would be appriciated.
> 
>   Thanks
> 
> -- Anant
>    anantk at yahoo.com
> 
> 
> Sent via Deja.com http://www.deja.com/
> Before you buy.
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 



From reic0024 at ub.d.umn.edu  Mon Nov  8 12:00:39 1999
From: reic0024 at ub.d.umn.edu (Aaron J Reichow)
Date: Mon, 8 Nov 1999 11:00:39 -0600
Subject: Fastest GUI toolkit?
In-Reply-To: <01bf29f1$cbfd5e60$5a41c3d1@Cyberus.cyberus.ca>
References: <01bf29f1$cbfd5e60$5a41c3d1@Cyberus.cyberus.ca>
Message-ID: <Pine.SO4.4.02.9911081059320.5258-100000@ub.d.umn.edu>

> I would like to draw 1000 lines in 1/10 of a second on a P166 machine.  (Am
> I way off the mark??)  I was hoping that someone would have tried different
> packages and could give me pointers in the right direction.

Perhaps wxWindows/wxPython?  I've not done much work with it, but I would
suppose that it is a bit faster than tkinter, as the GUI routines are
written in C++ and not Tcl/Tk.  Am I right?

Aaron




From greg.ewing at compaq.com  Wed Nov 24 05:46:32 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Wed, 24 Nov 1999 23:46:32 +1300
Subject: automatic "lotjes trekken"
References: <81cqnv$2ip$1@vvs.superst.iae.nl> <81e3s2$j14$2@newshost.accu.uu.nl> <81e41d$j14$3@newshost.accu.uu.nl>
Message-ID: <383BC208.4795B8CB@compaq.com>

Martijn Faassen wrote:
> 
> So,
> each time a person is drawn by someone, the program removes that program
> from everybody else's lists (if that person was in there in the first
> place).

That doesn't guarantee that everyone gets drawn, though.
You might get down to drawing for the last person and find
that there's nobody left on their list, even though there
are people remaining on other lists that have never been
drawn.

I can't think of any method that wouldn't have to backtrack
under some circumstances.

But is that really a problem in practice? How big a data
set are we talking about? Unless you have a REALLY BIG
family with lots and lots of conflicts, I'd expect a
solution to be found fairly quickly in most cases, even
if theoretically it could take a long time for some
pathological cases.

Greg



From rvollmert at gmx.net  Tue Nov 16 16:14:03 1999
From: rvollmert at gmx.net (Robert Vollmert)
Date: Tue, 16 Nov 1999 22:14:03 +0100
Subject: list.without()?
In-Reply-To: <3831BE57.366ED454@appliedbiometrics.com>; from tismer@appliedbiometrics.com on Tue, Nov 16, 1999 at 09:28:07PM +0100
References: <NDBBIAPHEKGOBIHCBLIIEEOOCAAA.mcfletch@vrtelecom.com> <3831BE57.366ED454@appliedbiometrics.com>
Message-ID: <19991116221403.D7366@krikkit.vollmert.home>

On Tue, Nov 16, 1999 at 09:28:07PM +0100, Christian Tismer wrote:
> Oh well, I overlooked that you have the try inside the while.
> That's bad. I thought I had read this:
> 
> def without( source, element):
>         temp = source[:]
>         try:
>                 while temp:
>                         temp.remove( element )
>         except:
>             return temp
>         return temp
> 
> Did you try this (I didn't, have no time).

Would it make a big difference if
  while temp:
were replaced by
  while 1:
?

-- 
Robert Vollmert                                      rvollmert at gmx.net



From hildeb at www.stahl.bau.tu-bs.de  Mon Nov 29 09:42:32 1999
From: hildeb at www.stahl.bau.tu-bs.de (Ralf Hildebrandt)
Date: 29 Nov 1999 14:42:32 GMT
Subject: fcntl/ioctl and O_NONBLOCK
References: <slrn8405je.q0.hildeb@stahlw06.stahl.bau.tu-bs.de> <slrn840205.sn4.hildeb@stahlw06.stahl.bau.tu-bs.de> <slrn83t5mv.bhk.hildeb@stahlw06.stahl.bau.tu-bs.de> <81ml5t$eer$0@216.39.151.169> <81p32a$m14$1@216.39.151.169> <81qj26$be6$0@216.39.151.169>
Message-ID: <slrn84546o.8s6.hildeb@stahlw06.stahl.bau.tu-bs.de>

On 28 Nov 1999 06:45:26 GMT, Donn Cave <donn at oz.net> wrote:

>Yes - only a guess, but it bites everyone at least once when trying
>to talk to another process via pipes.  After each write(), where your
>Python program is writing to the ``stdin'' of the plot process, add
>a flush(), i.e. stdin.flush().
>
>Python file objects use C buffered I/O, and when output is a pipe
>device it's block buffered.  This means your data doesn't go directly
>to the pipe, but into a buffer, and the pipe doesn't see it until a
>certain amount has accumulated.  The flush() forces all the buffered
>data to be written immediately anyway.  You can also control this
>by the buffer size parameter to the function that creates the file
>object, or just avoid a file object altogether and write directly
>to the pipe file descriptor.

Meanwhile I found out:

1) I use select() to see if there's s.th. to read/write
2) I use os.read(fd, buffersize) to read and make the fd non-blocking using 

        fcntl.fcntl(stdout.fileno(), FCNTL.F_SETFD, FCNTL.O_NDELAY)
        fcntl.fcntl(stderr.fileno(), FCNTL.F_SETFD, FCNTL.O_NDELAY)

The problem was that the normal commands std(out|err).read() are
blocking, because they don't encounter an EOF. They just keep reading.




From bjorn at roguewave.com  Mon Nov 29 11:47:22 1999
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Mon, 29 Nov 1999 09:47:22 -0700
Subject: Unbuffered stdout
Message-ID: <51C9F3C392D0D111A15600A0C99E64A9AF568E@localhost>

> leon <leonar_n at epita.fr> wrote:
> > I'm developping an IRC like software which must work on the 
> Web. I can't use
> > Java, and I'm using Python to program the CGI. But, I have 
> a big problem : I
> > don't know how to make the stdout unbuffered.
> > I tried to make some
> >     sys.stdout.write("test")
> >     sys.stdout.flush()
> > but it don't work.
> 
> looks fine to me.  maybe something else
> is wrong?
> 
> anyway, you could try running the inter-
> preter with the -u option:
> 
> $ python -?
> Unknown option: -?
> usage: python [option] ... [-c cmd | file | -] [arg] ...
> Options and arguments (and corresponding environment variables):
> ...
> -u     : unbuffered binary stdout and stderr (also PYTHONUNBUFFERED=x)
> ...
> 
> </F>

Note that some webservers automatically buffer cgi output until the cgi
closes stdout.  To the best of my knowledge there are no pretty workarounds
for this...  You should check your webserver's docs to see if this is the
case though.

--bjorn



From greg.ewing at compaq.com  Wed Nov 24 04:08:20 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Wed, 24 Nov 1999 22:08:20 +1300
Subject: rlcompleter
References: <383A87F6.5D2F@cs.bham.ac.uk>
Message-ID: <383BAB04.63DFCA4A@compaq.com>

Stuart Reynolds wrote:
> 
>   readline.parse_and_bind("\C-tab: complete")
> 
> should bind complete to Control+tab. However, this still seems to leave
> it bound to just tab.

Since ascii tab is already a control character, I doubt
whether it can tell the difference between Tab and
Control-Tab.

Greg



From tim_one at email.msn.com  Wed Nov 24 15:18:23 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 24 Nov 1999 15:18:23 -0500
Subject: Call for Programs: the "phonecode" benchmark
In-Reply-To: <80tro1$e7e$4@news.rz.uni-karlsruhe.de>
Message-ID: <000201bf36b9$0f0b1f20$292d153f@tim>

[Lutz Prechelt]
>  - Out of the following set of scripting languages:
>      Perl, Python, Rexx, Tcl
>    do you prefer one over the others?

Yes.

>  - Do you think one is inherently better than the others?

Yes -- but which one depends on what you're doing <0.5 wink>.  Suppose the
task is this:  next Tuesday, you will be assigned a problem to solve via
program, drawn at random from the set of all problems that have ever been
solved, or might plausibly be solved, via program.  You must choose the
programming language you're going to use next Monday, before knowing which
problem you must solve.  You must solve the problem by next Wednesday, else
your failure will be announced on Usenet and you'll suffer 10% more flames
than usual.

There are only two sane choices:  Python -- or suicide <wink>.

> ...
> I have prepared an interesting small programming problem
> [at http://wwwipd.ira.uka.de/~prechelt/phonecode/]

This was fun!  I enjoyed it.  Excellent writeup of the problem.

> ...
> I would now like to extend this investigation to scripting languages:
>
>   o Are differences between implementations using the same language
>     smaller or even larger than for "standard" programming languages?

My guess is larger:  "high level" data structures and algorithms are so much
easier to experiment with in high-level languages that a greater diversity
of approaches is natural.

>   o Do systematic differences between different scripting languages
>     exist?

Yes.  For example, not all scripting languages use "#" as a comment
character <wink>.

>   o How do the scripting languages compare to standard programming
>     languages such as C or Java?

Shorter code, easier to read, easier to write, easier to test, easier to
debug, less likely correct on first run (due to lack of static
declarations), more likely correct on last run (due to extensive
error-checking in the languages' runtime support -- Perl has to be run
with -w to get this, though), slower to run an equivalent algorithm, but
more likely to come up with a better algorithm.

> ...
> The effort required is typically somewhere between 1 and 8 hours.

Goodness -- that's between a month and a year's work for Fredrik <wink>.

cheerleadingly y'rs  - tim





From mstenber at cc.Helsinki.FI  Thu Nov 25 05:17:58 1999
From: mstenber at cc.Helsinki.FI (Markus Stenberg)
Date: 25 Nov 1999 12:17:58 +0200
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <867lj85g4c.fsf@g.local> <al8yabo9fm3.fsf@sirppi.helsinki.fi> <86so1v4jlj.fsf@g.local>
Message-ID: <al8hfiaan8p.fsf@sirppi.helsinki.fi>

Gareth McCaughan <Gareth.McCaughan at pobox.com> writes:

.. snip..

> Dylan is sort of nice. I can't think of any purpose for which
> I'd prefer Dylan to Common Lisp, though; the strengths of the
> two languages are in similar places, and I think CL wins.

I have always found lisp-ish syntax of "masses of ()s" disconcerting in
large amounts - admittedly, I've written more than few lines of
[ce]lisp/scheme/.., but never enjoyed browsing the results very much. I
consider Dylan to be "readable" clisp.

Admittedly, if the notation of clisp doesn't hinder you, it's _very_
powerful tool - I just can't wrap my head around larger clisp constructs..

>From trivial example - do you consider (+ 2 (* 3 4)) more readable than 2 +
3 * 4 most languages (including Dylan) allow me to do?

> -- 
> Gareth McCaughan  Gareth.McCaughan at pobox.com
> sig under construction

-Markus

-- 
	"If you want to travel around the world and be invited to speak at
	a lot of different places, just write a Unix operating system."
		-- Linus Torvalds



From robin at jessikat.demon.co.uk  Tue Nov  2 11:37:06 1999
From: robin at jessikat.demon.co.uk (Robin Becker)
Date: Tue, 2 Nov 1999 16:37:06 +0000
Subject: Assorted Questions regarding the future of Python...
References: <6D8A17398E28D3119F860090274DD7DB49833A@pces.cadlab.it>
 <m3bt9cgbwy.fsf@phaedrus.zen.xom>
Message-ID: <flG2+NAyMxH4Ewxf@jessikat.demon.co.uk>

In article <m3bt9cgbwy.fsf at phaedrus.zen.xom>, Rob Hodges
<s323140 at student.uq.edu.au> writes
>Alessandro Bottoni <Alessandro.Bottoni at think3.com> writes:
>
>> 3) Visual Basic has two advantages on Python: it can compile programs (OK: I
>> know that we can generate "frozen" Python programs with "Freeze") 
>
>Unless something has changed dramatically since I last used windoze,
>VB is not compiled (in the sense of native machine code) any more than
>a frozen python program; rather, VB "executables" have a small header
>of native code that launches the vbrunxxx.dll (interpreter) to run the
>bytecode that makes up the rest of the file.  Which is more or less
>the equivalent of writing "#! /usr/bin/env python" at the top of your
>python script in UN*X, providing a corresponding .pyc has been
>generated.  Or double-clicking a .pyc file in windoze, for that
>matter.  So either way, your users have to either have the python
>interpreter or the appropriate vb dll somewhere on their system.
>
>Someone please correct me if I'm wrong.  (Though there's no way I'll
>ever touch VB <*shudder*>; just that I wouldn't like to be spreading
>disinformation.)
>
>-Rob
>
there's a pretty easy way to tell; just implement a null loop and time
it. Loops in VBA, VB script et al. are pretty slow.
-- 
Robin Becker



From fdrake at acm.org  Tue Nov 16 09:54:10 1999
From: fdrake at acm.org (Fred L. Drake, Jr.)
Date: Tue, 16 Nov 1999 09:54:10 -0500 (EST)
Subject: docs
In-Reply-To: <000a01bf2ffc$3e70b0e0$042d153f@tim>
References: <14380.25301.616025.322479@weyr.cnri.reston.va.us>
	<000a01bf2ffc$3e70b0e0$042d153f@tim>
Message-ID: <14385.28690.800102.134431@weyr.cnri.reston.va.us>

Tim Peters writes:
 > Ask MS for the source code <wink>.  Short of that, there's no reason to
 > suppose it's a problem with the docs, and every reason to suspect the
 > obvious.

  The catch is that no matter how obvious it is, I don't want to lock
people out of the docs by requiring they update to a newer service
pack; patching an OS is never trivial (no matter how easy the
installation of the patches), and prone to break things (esp. for NT
service packs!).  I'm inclined to change the generation of the comment 
in the HTML simply to avoid having this problem.  It's annoying, but
easily handled.  (And people who read the CVS checkin messages may
have noticed that I've already handled it that way. ;)


  -Fred

--
Fred L. Drake, Jr.	     <fdrake at acm.org>
Corporation for National Research Initiatives



From Gaetan_Corneau at baan.com  Wed Nov  3 10:27:00 1999
From: Gaetan_Corneau at baan.com (Gaetan Corneau)
Date: Wed, 3 Nov 1999 10:27:00 -0500 
Subject: C++ embeding API?
Message-ID: <816010E2456BD111A48700805FBBE2EEFDF74D@ex-quebec-u1.baan.com>

Hello,

Is anyone working on a high-level C++ embedding API?
I would be interested in that, and could help any ongoing effort in that
direction.
I have no experience writing C++ apps with embedded Python (I'm just
beginning to read the doc), but I am an experienced C++ programmer.

Thanks,
______________________________________________________
   Gaetan Corneau
   Software Developer (Quality Assurance Team)
   BaaN  Supply Chain Solutions  
   E-mail: Gaetan_Corneau at baan.com        
   Compuserve: Gaetan_Corneau at compuserve.com  
   ICQ Number: 7395494             
   Tel: (418) 266-8252          
______________________________________________________
"Profanity is the one language all programmers know best"




From huiray at gate.sinica.edu.tw  Mon Nov 29 13:33:59 1999
From: huiray at gate.sinica.edu.tw (SINICA)
Date: Tue, 30 Nov 1999 02:33:59 +0800
Subject: Access violation while trying to replace stdout with a local disk file
References: <1268224356-30441862@hypernet.com>
Message-ID: <000a01bf3a98$4f13c940$36f96d8c@sylvia>

> They both need to be linked to the same c-runtime library, (ie,
> "Multithreaded DLL" in Settings | Compile | Code Generation).
>
> - Gordon
>
>

i've changed the c-runtime to "Multithreaded " for myprog.exe
and "Multithreaded DLL" for python15.dll
but it still refuse to work,
shall i change to "Multithreaded DLL" instead of "Multithreaded" for
myprog.exe ?

anyway, thank for your quick response.

yap




From da at ski.org  Fri Nov 26 13:28:34 1999
From: da at ski.org (David Ascher)
Date: Fri, 26 Nov 1999 10:28:34 -0800 (Pacific Standard Time)
Subject: Installing NumPy in a non-standard place...
In-Reply-To: <y0jhfi9xkxz.fsf@vier.idi.ntnu.no>
Message-ID: <Pine.WNT.4.04.9911261022440.198-100000@rigoletto.ski.org>

On 26 Nov 1999, Magnus L. Hetland wrote:

> Installing Python is a dream - but NumPy seems to be a bit stubborn...
> How can I tell it where I want to install it? I have copied
> Makefile.pre.in from my python installation, but still it wants to
> install it in a standard location (that I don't have access to,
> because I don't have root access on the machine)... Should I just copy
> all the files manually? Am I missing something? (Couldn't see anything
> in the installation instructions...)

Take installthis.py and add early:

import sys
sys.exec_prefix = '/wher/ever'

The long-term strategy has to be to get distutils-sig to replace the tools
we built for NumPy.

--david




From akoch at rbg.informatik.tu-darmstadt.de  Tue Nov 16 08:11:02 1999
From: akoch at rbg.informatik.tu-darmstadt.de (Andreas Koch)
Date: Tue, 16 Nov 1999 14:11:02 +0100
Subject: Integrate Phyton in a (Delphi) App ....
Message-ID: <383157E6.48E8@rbg.informatik.tu-darmstadt.de>

Hi , i plan writing a huge app , in Delphi (Win95).
It will have lots of classes for the various game
objects , and i currently think about writing this
class tree in Phython , to allow users changing it
and adding new stuff.

(for short , i call the delphi app winapp )

So :
1-How much work is it to integrate Phyton into a win app ?
  The app will hold the complete framework for user
  interface , movement , graphics , the phyton classes need to
  call and be called by my procedures
2-How fast is Phyton ? Speed is essential , i will have 
  at least 100 objects per frame that need to be called and
  (usually) call my winapps "drawtile" function.
3-Is it possible to restrict certain program parts
  from using certain code parts ? For example , don't
  allow while-loops in a user definable module ?
4-Is it possible for my winapp to execute some of
  the phyton scripts in single step mode ?
 
							thanks , Andreas



From bwarsaw at cnri.reston.va.us  Sat Nov  6 00:15:06 1999
From: bwarsaw at cnri.reston.va.us (Barry A. Warsaw)
Date: Sat, 6 Nov 1999 00:15:06 -0500 (EST)
Subject: Announcing Mailman 1.1
Message-ID: <14371.47450.186806.711568@anthem.cnri.reston.va.us>

I have just uploaded Mailman version 1.1 to www.list.org, and this
will hopefully soon be mirrored to ftp.gnu.org.  This release is
precipitated by the removal of all GIFs from the distribution.  There
have also been some performance improvements, although lots more can
still be done. ;)

See the NEWS file excerpt below for details of changes.  Sadly, the
internationalization effort has not yet been integrated (mostly due to
time constraints of the core maintainers).

Mailman, the GNU mailing list manager, is software to help manage
email discussion lists, much like Majordomo and Smartmail.  Unlike
most similar products, Mailman gives each mailing list a web page, and
allows users to subscribe, unsubscribe, etc. over the web.  Even the
list manager can administer his or her list entirely from the web.
Mailman also integrates most things people want to do with mailing
lists, including archiving, mail-to-news gateways, and so on.  Mailman
is writtein primarily in Python, with a little bit of C for security.

See http://www.list.org or
http://www.gnu.org/software/mailman/mailman.html for details and
downloading.

-Barry

-------------------- snip snip --------------------
- All GIFs removed.  See http://www.gnu.org/philosophy/gif.html for
  the reason why.

- Improvements to the Pipermail archiver which make things faster.
  Primary change is that the .txt files are not gzip'd on every posted
  message.  Instead, use the new cron script `nightly_gzip' to gzip
  the .txt file in batches (this means that the .txt file will lag
  behind the on-line archives a little).

- From the C drivers programs, Python is invoked with the -S option.
  This tells Python to avoid importing the site module, which can
  improve start up time of the Python process considerably.  Note that
  the command line script invocation has not been changed.

- New configuration variables PUBLIC_EXTERNAL_ARCHIVER and
  PRIVATE_EXTERNAL_ARCHIVER which can contain a shell command string
  for os.popen().  This can be used to invoke an external archiver
  instead of the bundled Pipermail archiver.  See Defaults.py for
  details.

- new script `bin/find_member' which can be used to search for a
  member by regular expression.

- More child processes are reaped, which should eliminate most
  occurances of zombie processes.

- A few small miscellaneous bug fixes (including PR#99, PR#107) and
  improvements to the file locking algorithms.
-------------------- snip snip --------------------

<p><a href="http://www.list.org/">Mailman 1.1</a> - the GNU mailing
list manager, implemented primarily in Python.  (05-Nov-1999)



From T.Middletin at news.vex.net  Fri Nov 19 05:23:16 1999
From: T.Middletin at news.vex.net (T. Middletin)
Date: 19 Nov 1999 10:23:16 GMT
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <JlFX3.3659$s87.67930@dfiatx1-snr1.gtei.net> <80oqn9$ocq$2@newshost.accu.uu.nl> <oLUX3.286$yA4.4821@dfiatx1-snr1.gtei.net> <80p9su$ppo$2@newshost.accu.uu.nl> <80piru$13rl$1@hub.org> <80re5d$55a$1@newshost.accu.uu.nl> <fjPY3.770$An3.14793@newsfeed.avtel.net> <810jto$lo0$1@newshost.accu.uu.nl> <811pdu$dvo$1@newshost.accu.uu.nl>
Message-ID: <8138ek$22mj$1@hub.org>

On 18 Nov 1999 21:00:46 GMT, Martijn Faassen <m.faassen at vet.uu.nl> wrote:
>with the same speech patterns _from vex.net_, I meant! deja nor I are that

Argh! They're on to me. I see i'll have to randomise news servers 
for posting more, and deja too. (-; Oh yaw, and type with a dutch
accent! Happy trails.
...





From robin at alldunn.com  Mon Nov 15 15:55:58 1999
From: robin at alldunn.com (Robin Dunn)
Date: Mon, 15 Nov 1999 12:55:58 -0800
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <14384.16479.836502.161209@anthem.cnri.reston.va.us> <80pifh$12n1$1@hub.org>
Message-ID: <8s_X3.676$pw2.12189912@news.randori.com>

>
> I must say i'm a little worried about bandwidth issues after the
> discussions here. But on the other hand... we'll never know what the
> system can or can't take until we try right? Heh.
>
> Or perhaps it would be prudent for me to look into possibilities
> further.... i'm in a bit of a quandry. I've gotten more response than
> I expected (which has been nice... but a little frightening too!).
>

Could it be moved to the Starship or perhaps to a machine at CNRI to be part
of python.org?


--
Robin Dunn
Software Craftsman
robin at AllDunn.com
http://AllDunn.com/robin/
http://AllDunn.com/wxPython/  Check it out!






From claudius at catlover.com  Sat Nov  6 12:08:23 1999
From: claudius at catlover.com (claudius at catlover.com)
Date: Sat, 06 Nov 1999 17:08:23 GMT
Subject: type object of an instance without instantiation?
Message-ID: <bgZU3.19722$Ec1.199882@news1.frmt1.sfba.home.com>

I have the following code (well, in summary):

class BigClass:
	def __init__(self):
		...does some horrifyingly nasty operation...

def some_generic_fn(object):
	if hasattr(object, '__class__') and object.__class__ == BigClass:
		...do some BigClass-specific thing...

Is there a way to create a type object for the instance of a BigClass object
without actually creating an instance of one?



From mwh21 at cam.ac.uk  Thu Nov 11 10:02:24 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 11 Nov 1999 15:02:24 +0000
Subject: print "foo", without a space
References: <382AD3BB.69899CBD@helsinki.xx>
Message-ID: <m3u2mtoyv3.fsf@atrus.jesus.cam.ac.uk>

Mikael Johansson <mikael.johansson at helsinki.xx> writes:

> Hello All!
> 
> Is there another way of getting print to not make a newline than
> using:
>     print "foo",  ?

I think there's some way invloling sys.stdout.softspace, but I'm no
sure what. However...

> This because the comma always produces a space which necessary
> isn't that nice, for example:
> 
> for i in range (1,30):
>     print i,
>     if i%10==1 and i<>11: print "st",
>     elif i%10==2 and i<>12: print "nd",
>     elif i%10==3 and i<>13: print "rd",
>     else: print "th",
>     print "fish"
>  
> Now you would like the "endings" to come right after the number.
> There is naturally always some way to get around this, in this
> case for example:
> 
> end=["st","nd","rd","th"]
> for i in range(1,30):
>     print `i`+end[max(min(abs(i)%10,4)-1-
>     4*(not(int('0'+`abs(i)`[-2:-1])-1)),-1)],'fish'
> 
> OK, this got a bit out of hand, 

!

> the above looks terrible, does 
> anyone see how it could be simplified? At least it's more general
> than the above as it can handle the whole range of integers,
> including negative ones :-) However it shows that it would be
> less obfuscating to have a print statement that could continue
> printing on the same line without a space.

Do you know about the string % operator? Try:

def ordinate(n):
    if n%10 == 1 and n <> 11:
        return "st"
    if n%10 == 2 and n <> 12:
        return "nd"
    else:
        return "th"

print "This is the %d%s fish"%(i,ordinate(i))

It's generally the way to go for formatting output.

> If there isn't an easy way for this, I suggest the following:
>     print "foo".
> So that a dot at the end of the line would mean "no space either".

With all due respect: no chance.
 
> Appreciating any comments.

Hoping this helps,
Michael



From fredrik at pythonware.com  Wed Nov 24 06:37:22 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 24 Nov 1999 12:37:22 +0100
Subject: Python complaints
References: <XFMail.991124113221.mikael@isy.liu.se> <383BC9C2.901A2D6@python.net>
Message-ID: <018401bf3670$46b24af0$f29b12c2@secret.pythonware.com>

Thomas A. Bryan <tbryan at python.net> wrote:
> I generally don't mind this aspect of Python, but I've recently found that 
> it can cause problems when transmitting code.  For example, I saw some 
> code that I wanted to copy in DejaNews c.l.py archives, but the lines 
> wrap in narrow columns in DejaNews.  I had to reindent the code.  Grrr.

but you didn't have to do that, really -- dejanews has a
small "view original Usenet format" link at the bottom of
each message.

(or you can use the "mail to a friend" link to mail your-
self an exact copy of the original message...)

</F>

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





From phil at ricochet.net  Tue Nov 30 02:38:49 1999
From: phil at ricochet.net (phil)
Date: Tue, 30 Nov 1999 07:38:49 GMT
Subject: PyArg_ParseTuple() and oprtional remaining argument lists
References: <000c01bf3a70$ae184ec0$3acbd9c2@peridot.optichrome.com>
Message-ID: <38437a53.7776331@news.ricochet.net>

On Mon, 29 Nov 1999 13:50:22 -0000, "Adrian Eyre"
<a.eyre at optichrome.com> wrote:

Thanks, that fixed both the problems.

phil.

>Aha. That's the clue. Make sure that your PyMethodDef struct is like
>this:
>
>static struct PyMethodDef methods[] = {
>    { "method1",   method1, 1 },
>    { "method2",   method2, 1 },
>    { NULL,        NULL,    0 } 
>};
>
>Not like this:
>
>static struct PyMethodDef methods[] = {
>    { "method1",   method1 },
>    { "method2",   method2 },
>    { NULL,        NULL    } 
>};
>
>--------------------------------------------
>Adrian Eyre <mailto:a.eyre at optichrome.com>
>Optichrome Computer Solutions Ltd
>Maybury Road, Woking, Surrey, GU21 5HX, UK
>Tel: +44 1483 740 233  Fax: +44 1483 760 644
>http://www.optichrome.com 
>--------------------------------------------
>
>
>

The world is divided into two sorts of people: those that think the
world is divided into two sorts of people and those that don't.



From gmcm at hypernet.com  Sat Nov  6 20:58:57 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Sat, 6 Nov 1999 20:58:57 -0500
Subject: exec / execfile namespace problem
In-Reply-To: <000b8d9b.b9d3bdca@usw-ex0101-003.remarq.com>
Message-ID: <1270182562-5943662@hypernet.com>

Karan Vasudeva has problems with exec:

> .... my
> problem is, when i call my script from another script (by doing
> an exec, thats what i need to do, you see), the interpreter stops
> at httplib.py as it can't find that socket called self.sock .
> This appears to be a namespace issue. I just want to know how to
> do this right.

exec is pretty much only used when you don't know what 
you'll be running at the time you wrote it. If you know what 
you're running at the time you write it, there's no reason to use 
exec. You probably just want:

  import cool

and maybe:
  cool.main(args)

Many python scripts have the form:
-----------------
def main():
  # whatever

def otherthing():
  # stuff used by main, or usable by others

if __name__ == '__main__':
  main() # if invoked as a script, run main()
---------------------------------
 
Then they can run as the top level script, or they can be 
imported, and any functions, classes etc. used by the 
importer.



- Gordon



From framiere at netcom-service.com  Mon Nov 15 06:12:59 1999
From: framiere at netcom-service.com (Florent Ramière)
Date: Mon, 15 Nov 1999 12:12:59 +0100
Subject: Python ISAPI
Message-ID: <80oq15$ln8$1@oceanite.cybercable.fr>

Hello,

    i am looking for an python ISAPI, but could not find any.
    Could you tell me where i can get such a goodie ?
    Does anybody ever worked on this ?

Thanks.





From steven_shaw at adc.com  Thu Nov  4 01:57:22 1999
From: steven_shaw at adc.com (steven_shaw at adc.com)
Date: Thu, 04 Nov 1999 06:57:22 GMT
Subject: SWIG and function name question
References: <7vmc7e$8h$1@marcie.fissiontech.com>
Message-ID: <7vraoh$6tm$1@nnrp1.deja.com>

You won't get them with your 'from module import *' because they
begin with underscore. Names beginning with underscore are "private"
by convention in Python. The 'from module import *' implements this
"convention". It is the only for of access restriction in Python.

Probably someone else can explain it better.

-SS

In article <7vmc7e$8h$1 at marcie.fissiontech.com>,
  kain at dynup.net wrote:
> Hey folks,
>
> I downloaded and started to learn SWIG today and I ran into a little
problem.
> In the C code there is a structure like this:
>
> struct __foo { ... };
>
> When I import my SWIG created module using 'import mymodule' I have
access
> to the struct's set/get operation via mymodule.__foo_set_member() but
if I
> import my module using 'from mymodule import *' I have access to all
of my
> modules functions accept the two get/set functions that begin with __
> (I'm using dir() to check this).  This isn't too much of a problem
for me
> but I'm just wondering what's going on.  Can anyone clear this up for
me?
>
> Thanks,
> Scott
>
>


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



From mwh21 at cam.ac.uk  Wed Nov 10 02:56:07 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 10 Nov 1999 07:56:07 +0000
Subject: IDLE and rlcompleter [Was: Re: making the Python case to my dept.]
References: <000201bf2b3c$a977dc20$d32d153f@tim>
Message-ID: <m3yac6n5k8.fsf@atrus.jesus.cam.ac.uk>

"Tim Peters" <tim_one at email.msn.com> writes:

> [Michael Hudson]
> > ...
> > While we're on the subject of idle, is there any easy (preferably one
> > or two keystrokes) that way to import/reload the module being edited
> > into the Python shell and bring said shell to the front for instant
> > interactive play? That'd be ?bercool.
> 
> Already implemented, in the current CVS IDLE.

It's that bloody time machine again! I'd have sworn I looked before I
posted.

Regards,
Michael



From gmcm at hypernet.com  Fri Nov  5 21:47:56 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Fri, 5 Nov 1999 21:47:56 -0500
Subject: __str__ vs. __repr__
In-Reply-To: <7vvni7$19q0$1@nntp6.u.washington.edu>
Message-ID: <1270266026-923418@hypernet.com>

Donn Cave writes:

> Quoth "Tim Peters" <tim_one at email.msn.com>:
> 
> | Let me back off to what repr and str "should do":
> |
> | repr(obj) should return a string such that
> |     eval(repr(obj)) == obj

> For real?  I'm very used to seeing repr's that don't eval at all
> -
> 
> '<socket object, fd=3, family=2, type=1, protocol=0>'
> "<open file '<stdout>', mode 'w' at 80005660>"
> '<exceptions.NameError instance at 80048ae0>'

import StringIO
x = ['a', 1, 'b', {'x':(4,5L),'z':None}]
f = StringIO.StringIO()
f.write(repr(x))
y = eval(f.getvalue())
if y == x:
  print "You betcha"

I've even been known to implement __repr__ in classes and C 
extension types so this property holds true, and I've been 
bitten by the fact that UserList just forwards repr to self.data.

- Gordon



From godzilla at netmeg.net  Thu Nov  4 23:38:37 1999
From: godzilla at netmeg.net (Les Schaffer)
Date: Fri, 05 Nov 1999 04:38:37 GMT
Subject: ~ 1 = -2 ???
References: <m2vh7hk1ys.fsf@netmeg.net> <slrn824m9l.ngk.wtanksle@hawking.armored.net>
Message-ID: <m2so2lk08p.fsf@netmeg.net>

william tanksley added this marble to my collection:

> The marble about 32-bit twos' complement signed integers.  They're
> not four-bit, BTW.

its not even midnite and i am already brain-dead, yeeesh.

i was doing ops on 4 bit numbers, so i fooled myself.

i should have said:

0000 0001 --> 1111 1110 

and to keep only the 4 lower bits...

1111 1110 & 0000 1111 --> 0000 1110 

which equals 14, not 13 (sheesh... nobody is ever gonna listen to me
again)

so for my code, i will do

def not4bit(x):
      return ~x & 15

>>> not4bit(1)
14

ahh, that feels better....

goodnite world.

les schaffer



From darcy at vex.net  Thu Nov 11 15:44:43 1999
From: darcy at vex.net (D'Arcy J.M. Cain)
Date: 11 Nov 1999 20:44:43 GMT
Subject: ASCII delimited files
References: <942231502.1914033961@news.interbusiness.it> <80ekhb$1qrh$1@hub.org> <80et0b$2df8$1@hub.org> <80evr8$2lc7$1@hub.org> <80f11i$2nof$1@hub.org>
Message-ID: <80f9rr$9m0$1@hub.org>

D'Arcy J.M. Cain <darcy at vex.net> wrote:
> OK, this one does the right thing.

*sigh*

Except that I accidently posted an earlier test version.

> 	SEP='|'

Should be SEP = '\0'

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



From da at ski.org  Thu Nov 11 12:22:30 1999
From: da at ski.org (David Ascher)
Date: Thu, 11 Nov 1999 09:22:30 -0800 (Pacific Standard Time)
Subject: method overriding trick
In-Reply-To: <14378.62498.649973.150075@bitdiddle.cnri.reston.va.us>
Message-ID: <Pine.WNT.4.04.9911110920120.207-100000@rigoletto.ski.org>

On Thu, 11 Nov 1999, Jeremy Hylton wrote:

> Overriding methods of a base class is a little clunky in Python.  You
> need to explicitly name the base class that implements the method, and
> you call its as an unbound method, passing self as the first argument.
> Example: Base.aMethod(self, arg)
> 
> I was reading about MzScheme the other day (a Scheme variant with
> object and interface support), and noticed that the MzScheme approach
> translates well to Python.  MzScheme class defintions allow a rename
> declarations that create a name in the class being defined that is
> bound to a name in a superclass.  In Python, rename is just called
> assignment. 
> 
> class Base:
>     def aMethod(self, arg):
>         return arg * 2
> 
> class Derived1(Base):
>     super_aMethod = Base.aMethod
> 
>     def aMethod(self, arg):
>         return self.super_aMethod(arg) + 1
> 
> I haven't seen this strategy for overriding mentioned before, although
> I did not search the list archives.  When I described it to Guido, he
> said, "Does that work?"  Has anyone used this trick before?
> 
> It doesn't solve the must-name-the-base-class problem, but it helps
> manage it better.  If you keep the override assignments at the top of
> the class statement, then you only have one place to look in each
> class when the inheritance hierarchy changes.

Cute!  You can also make it a mixin, along the lines of:

import new
class Mixin:
    def __getattr__(self, k):
	if k[:len('super_')] == 'super_':
	    for b in self.__class__.__bases__:
		m = getattr(b, k[len('super_'):], None)
		if m: 
		    return new.instancemethod(m, self, self.__class__)
	    else:
		raise AttributeError, k
	else:
	    raise AttributeError, k

class Base(Mixin):
    def f(self): 
	print 'Base.f'
    
class Sub(Base):
    def f(self):
	print 'Sub.f'
	self.super_f()


s = Sub()
s.f()


Not especially fast, of course, but doesn't require as much human memory
load. 

--david




From mal at lemburg.com  Fri Nov 12 14:02:08 1999
From: mal at lemburg.com (M.-A. Lemburg)
Date: Fri, 12 Nov 1999 20:02:08 +0100
Subject: __str__ vs. __repr__
References: <000001bf275c$02370d20$1d2d153f@tim> <7vvni7$19q0$1@nntp6.u.washington.edu> <382BEC79.EC65913D@maxtal.com.au>
Message-ID: <382C6430.8DC07E3E@lemburg.com>

skaller wrote:
> ...viperi's version of repr...
> You may be wondering how this works. The answer is that Viperi does not
> use bytecode, it executes the 'Abstract Syntax Tree' directly.
> Consequently, the original code can be recovered. (without comments,
> or the original formatting, of course).

Could you point me to some resources ? Using ASTs for execution
is an interesting subject and I would like to know how you deal
with Python dynamic nature in this context (are the ASTs self
modifying ?).

BTW, all your software links are broken...

Thanks,
-- 
Marc-Andre Lemburg
______________________________________________________________________
Y2000:                                                    49 days left
Business:                                      http://www.lemburg.com/
Python Pages:                           http://www.lemburg.com/python/



From skip at mojam.com  Mon Nov 29 15:39:45 1999
From: skip at mojam.com (Skip Montanaro)
Date: Mon, 29 Nov 1999 14:39:45 -0600 (CST)
Subject: Python bugs
In-Reply-To: <001101bf3a8f$5fe300f0$3acbd9c2@peridot.optichrome.com>
References: <51C9F3C392D0D111A15600A0C99E64A9AF5690@localhost>
	<001101bf3a8f$5fe300f0$3acbd9c2@peridot.optichrome.com>
Message-ID: <14402.58513.737532.298175@dolphin.mojam.com>

    >>> s = shelve.open("foo")
    >>> l = []
    >>> s["bar"] = l
    >>> s["bar"]
    []
    >>> l.append("baz")
    >>> s["bar"]
    []

I think you're asking for magic that really can't happen (at least not
easily).  Shelve transparently does the mapping to strings that most/all
dbm-like packages require, so when you execute

    s["bar"] = l

you'd don't get the usual object-to-name binding you expect.  Instead, l is
marshalled into Pickle's string form and that string is stuffed into the
underlying dbm file keyed by the string "bar".  Similarly, when getting the
value of s["bar"], the read+unpickle operation is performed.

When you append "baz" to l, it has nothing to do with s["bar"].  While this
isn't as transparent as you'd like, the lack of transparency has one big
advantage: speed.  Because you're forced to explicitly read and write
elements of a shelve object, e.g.:

    s = shelve.open("foo")
    l = s["bar"]
    ... do lots of stuff with l ...
    s["bar"] = l

instead of

    s = shelve.open("foo")
    ... do lots of stuff with s["bar"] ...

the result is that all operations involving l are done in memory instead of
by constanting marshalling and unmarshalling a possibly very big item out of
a file.

All that aside, I believe the good folks at Digital Creations worked some
magic using ExtensionClasses a few years ago to make this work in certain
contexts.

Skip Montanaro | http://www.mojam.com/
skip at mojam.com | http://www.musi-cal.com/
847-971-7098   | Python: Programming the way Guido indented...



From mwh21 at cam.ac.uk  Mon Nov 29 12:38:25 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 29 Nov 1999 17:38:25 +0000
Subject: performance of tuple-less Python?
References: <HSx04.8911$cz3.321455@nnrp1.uunet.ca>
Message-ID: <m3d7stryem.fsf@atrus.jesus.cam.ac.uk>

"Greg Wilson" <gvwilson at nevex.com> writes:

> Has anyone ever built a version of Python that didn't include tuples, and
> looked at the performance of code that consistently used lists instead?  

I doubt it.

> I'm curious as to whether the use of tuples (which I believe are
> conceptually redundant) actually has a significant positive impact
> on the interpreter's performance on non-trivial codes (i.e. once
> caching, instruction ordering, and all other effects have come into
> play).

Tuples are immutable. This is important in some respects; they're
hashable for instance, and can be guaranteed not to contain cycles.

I don't think they're there for reasons of performance.

M.



From kern at temp1.caltech.edu  Wed Nov 10 14:50:45 1999
From: kern at temp1.caltech.edu (kern at temp1.caltech.edu)
Date: 10 Nov 1999 19:50:45 GMT
Subject: Using GNU C complier as the "default" windows complier
References: <4D0A23B3F74DD111ACCD00805F31D8100DB9151B@RED-MSG-50>
Message-ID: <80cial$gfn@gap.cco.caltech.edu>

In article <4D0A23B3F74DD111ACCD00805F31D8100DB9151B at RED-MSG-50>, Bill Tutt <billtut at microsoft.com> writes:
> cygwin isn't thread safe at this moment in time.
> If anything, Python should be compiled using mingw32 instead of cygwin.
> Esp, in order to get the Win32 extensions to be useful.
> 
> However.... I have no idea if the Win32 extensions actually can be compiled
> successfully using mingw32 as a substantial portion of them are C++, and
> even if they could be compiled, there would be binary incompatability
> problems due to using C++. (C++ mangling is different for all C++
> compilers...)

C extensions work fine when compiled with mingw32 as long as you patch the headers 
slightly.  I have compiled M.A. Lemburg's C++ mxCrypto extensions with mingw32, and 
they seem to work fine.  LLNL's CXX doesn't run the last time I tried, but this may
be a problem with the options I used and not a fundamental incompatibility.

I have not had the opportunity to try GCC 2.95.2 yet.

> So given those two things, and neither Mark, I, or anybody
> else atm seems to care enough to bother trying to see if it'll work. The
> binary compatability problem is the biggest problem though. If a MSVC
> compiled Python install can't make use of a mingw32 Win32 extension DLL (or
> vice versa) then even bothering with trying becomes counter productive.

MSVC Python can use mingw32 extensions if the extensions are compiled properly.
See http://starship.python.net/crew/kernr/mingw32/Notes.html for instructions.
Paul Sokolovsky has patched the Python interpreter to compile with mingw32.  The 
build process still needs work, but the base is there.  http://www.is.lg.ua/~paul/devel

> (unless your willing to invest time in submitting changes to the source code
> to remove the C++ mangling issues) i.e. Turn all exported C++ objects into
> COM objects or something so you can just use COM vtbls to communicate.
> Needless to say, thats a lot of work that I'd rather not do. (I expect Mark
> won't be to gung ho to do that work either... :) )
> 
> Bill

[piggybacking below]

> > -----Original Message-----
> > From: Werner Fouche [mailto:wfouche at mindspring.com]
> > Sent: Wednesday, November 10, 1999 3:59 AM
> > To: python-list at python.org
> > Subject: Re: Using GNU C complier as the "default" windows complier
> > 
> > 
> > 
> > "Mike Callahan" <mcalla at home.com> wrote in message
> > news:ma4W3.7794$Ur4.223420 at news.rdc1.tn.home.com...
> > > This a question from a newbie but I was thinking...why 
> > doesn't the default
> > > Python code for windows use the GNU C complier for windows 
> > instead of the
> > MS
> > > complier. Would it be possible to distribute the GNU 
> > package along with
> > the
> > > Python package and thus make it easier to compile, extend, and embed
> > Python,
> > > use freeze, etc? I know it would take some work, but it 
> > seems it would
> > make
> > > windows Python a lot more like unix Python. Just a thought.

Bad idea.  Cygwin is huge, and anyone who needs it can get it herself.  Cygwin
alone is not the best compiler to use since it goes through a POSIX emulation
layer that slows things down.  mingw32 would be better since it uses MSVCRT.dll 
as its C runtime library (when properly configured).  However, as Bill mentioned,
C++ issues still would have to be worked out.

Here's how things stand as I can see:

  1) mingw32 extensions run on MSVC Python.
  2) The Python interpreter can be compiled with mingw32.
  3) C++ support is still iffy.
  4) freeze will probably work with mingw32, but I've never heard of anyone trying.
  5) Embedding?  Who knows?
  6) Use Cygwin for tools and not the compiler.

> > I had the same idea on the weekend. So I installed Cygwin
> > (http://www.cygnus.com/cygwin) on my NT system and after running
> > "./configure" got Python to compile and run. However, I could 
> > not get Python
> > threads to work.

Cygwin proper is not threadsafe.  mingw32 can be.

Robert Kern
kern at caltech.edu



From godzilla at netmeg.net  Thu Nov  4 23:01:22 1999
From: godzilla at netmeg.net (Les Schaffer)
Date: Fri, 05 Nov 1999 04:01:22 GMT
Subject: ~ 1 = -2 ???
Message-ID: <m2vh7hk1ys.fsf@netmeg.net>

i am doing some bit operations, and dont understand the following:

>>> ~ 1
-2

huh???? i would have thought that bitwise invert of 1 would be

0001 --> 1110 = 13 (base 10)

what marble am i missing?

les schaffer



From andres at corrada.com  Sun Nov 21 12:36:39 1999
From: andres at corrada.com (Andres Corrada)
Date: Sun, 21 Nov 1999 12:36:39 -0500
Subject: Order of keyword arguments
References: <1268914352-35423797@hypernet.com>
Message-ID: <38382DA7.A03A2023@corrada.com>

Gordon McMillan wrote:
> 
> Ah, but that would be "alphabetical order", not "entry order",
> and not much use to anybody if it were.
> 

I disagree. I think you are confusing entry order with type of ordering.
entry order could be alphabetical but it could also be by time of entry,
for example. My point still stands: an ordered dictionary would be a
useful data structure to have.

> Python data structures that would suit your apparent desires:
>  [ (optionname, optionvalue), ... ]

Mike Fletcher suggested this also and it is the way I'm getting to eat
my cake and have it too :-)

------------------------------------------------------
Andres Corrada-Emmanuel   Email: andres at corrada.com
Owner                     http://www.corrada.com/mamey
Mamey                     Phone: (413) 587-9595
------------------------------------------------------



From tekhir at yahoo.com  Tue Nov 16 21:56:33 1999
From: tekhir at yahoo.com (Jeremy Misavage)
Date: Tue, 16 Nov 1999 21:56:33 -0500
Subject: How I got Python to Work with IIS 4.0 on NT
Message-ID: <38321961.540FFEFD@yahoo.com>

The instructions I found on python.org didn't work for me totally.  IIS knew I
had .py as a script type, but it wouldn't execute them correctly.  However, I
did get it to work and this is how:

1. In the Management Console select your web site
2. Pull up Properties (I had to right click)
3. Go to the Home directory tab
4. Click configure
5. Select the App Mappings tab

This is where you say that .py files are handled by "c:\Program
Files\Python\python.exe" (by default)

"c:\Program Files\Python\python.exe" -u %s %s

I've also been told to put "" around the %s sometime.  Like so:

"%s" "%s"

MS Personal Web Server however doesn't have a Management Console so registry
hacking is the only way to get it to work.



From x at domain.com  Sat Nov 27 18:57:32 1999
From: x at domain.com (Cove Schneider)
Date: 27 Nov 1999 23:57:32 GMT
Subject: SMTP - again ;)
References: <19991126080017.26482.qmail@hotmail.com>
Message-ID: <38406fec$0$222@nntp1.ba.best.com>

Ulf Engstr?m <b2blink at hotmail.com> wrote:
> I've gotten my prog to work now, only have some problems with our 
> server/firewall, but that's hardly an issue.

> Anyway, something else came up, when I enter a name in the header:
> "To: Ulf Engstr?m" the mail gets to the dest as:

> To: Ulf.Engstr?m at mb04.swip.net if I use mb04.swip.net as outgoing server, 
> eventhough the address is completely different...and the add before @ is not 
> correct either, since it's the Real Name..
> Is there a way to get rid of that? In the To-field of the mail I just want 
> the name, not @anywhere...

Er.. Well, your most likely confusing your MTA because there's no address
in the To field. Though I suppose you could always do something like:

To: Foo Bar <>

and just use a blank address.. that shouldn't really pose any problems..

Cove

(at nbn dot com)



From thamelry at vub.ac.be  Thu Nov 25 04:29:18 1999
From: thamelry at vub.ac.be (Thomas Hamelryck)
Date: 25 Nov 1999 09:29:18 GMT
Subject: Python complaints
References: <000601bf36be$f1bf5e80$292d153f@tim>
Message-ID: <81ivhe$12l$1@mach.vub.ac.be>

Tim Peters (tim_one at email.msn.com) wrote:
: There is one (& only one) area where Python's indentation gets in my way a
: bit:  conceptual blocks that can't be written as blocks.  These typically

Right. Another example occurs in OpenGL, eg.

glBegin(GL_POLYGON)
   glVertex2f(-0.5, -0.5)
   glVertex2f(-0.5, 0.5)
   glVertex2f(0.5, 0.5)
   glVertex2f(0.5, -0.5)
glEnd()

This makes the code more readable, but it's not allowed in Python. I've 
got used to the indentation rule in Python, and I'm not annoyed by it that
much anymore, but I still think it is a major mistake.

Cheers,

------------------------------------------------
Thomas Hamelryck//Free University Brussels (VUB)
Intitute for Molecular Biology// ULTR Department
Paardestraat 65//1640 Sint-Gensius-Rode//Belgium
------------------------------------------------



From marten at btk.utu.fi  Fri Nov 26 06:18:51 1999
From: marten at btk.utu.fi (marten at btk.utu.fi)
Date: 26 Nov 1999 13:18:51 +0200
Subject: Tkinter: closing app with active messagebox
Message-ID: <uogchfqlg.fsf@GARLIC.i-did-not-set--mail-host-address--so-shoot-me>

Hi

Im writing a small script that asks the user for some input and checks
it from a mysql server. If the user doesn't respond the app should
close after a  timeout period. Normally this works fine, but if there
is an active messagebox (tkMessageBox.showinfo), the app won't close
until the OK button in the messagebox is clicked.

My question is: How do I force an active messagebox to close? Can this
be done with the standard dialog boxes in tkMessageBox or do I have to
write my own dialog boxes?

Any help appreciated

Yours

Marten Hedman			Marten.Hedman at btk.utu.fi
Turku Centre for Biotechnology
Abo Academy University & Turku University
Turku, Finland



From fredrik at pythonware.com  Tue Nov 16 11:33:46 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 16 Nov 1999 17:33:46 +0100
Subject: Anyway to use file masks with os.listdir()?
References: <Pine.SO4.4.02.9911160925560.22506-100000@ub.d.umn.edu>
Message-ID: <008a01bf3050$5f819cf0$f29b12c2@secret.pythonware.com>

Aaron J Reichow <reic0024 at ub.d.umn.edu> wrote:
> Is there any way to pass a filemask to os.listdir()?  For example, to list
> only files with the extension .jpg.  Any help would be appreciated! :)

it's spelled "glob":

http://www.python.org/doc/current/lib/module-glob.html

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
(news: just posted an errata for the 'first printing') -->




From andy at robanal.demon.co.uk  Sun Nov  7 11:41:18 1999
From: andy at robanal.demon.co.uk (Andy Robinson)
Date: Sun, 07 Nov 1999 16:41:18 GMT
Subject: __str__ vs. __repr__
References: <000001bf275c$02370d20$1d2d153f@tim>
Message-ID: <38257118.2127148@news.demon.co.uk>

"Tim Peters" <tim_one at email.msn.com> wrote:

>> Should I really assume stdout is capable of displaying Latin-1?
>
>No.  But I don't grasp why you think you need to know *anything* about this.
>Until Unicode takes over the world, there's nothing you can do other than
>tell users the truth:  "most of" printable 7-bit ASCII displays the same way
>across the world, but outside of that *all* bets are off.  It will vary by
>all of OS, display device, displaying program, and user configuration
>choices.
>
Well said. I'm going to pass this message around to my team at work as
it encapsulates so many of the issues involved in internationalisation
(and if you think inputting Fran?ois is bad, try figuring out a way to
clean up Japanese names using vi on an English Solaris box).

I have found there are only three really sane points of reference when
dealing with encodings - anywhere between the gaps just causes
confusion all around:

Level 1: Code that is assumed to work as above (tab, newline and
32-127)
Level 2: Code that keeps all 256 values intact without understanding
the contents, like Python strings.
Level 3:  A fully general multi-byte encoding toolkit, where the
programmer is in explicit control of which encoding a string is in,
and has capable libraries which can convert from one to the other, and
can reason in advance about whether a particular data set can survive
a particular round-trip conversion, and knows the exact capabilities
of the fonts ultimately used for display or printing.

AFAIK (3) does not exist yet - we got most of the way at work with
Unilib and a load of ad hoc Python code, and Java goes most of the
way.   I think the key concept is a special kind of string which is
tagged to know which encoding it is in.

Regards,

Andy





From tre17 at student.canterbury.ac.nz  Sun Nov 14 20:28:41 1999
From: tre17 at student.canterbury.ac.nz (Tim Evans)
Date: 15 Nov 1999 14:28:41 +1300
Subject: Other ways to disable Wm in Tkinter?
References: <80nabr$q62$1@nnrp1.deja.com>
Message-ID: <c757ljk1r1y.fsf@pc142.cosc.canterbury.ac.nz>

brian at liaone.ne.mediaone.net writes:

> Hello,
> I have a question on how to dynamically activate/deactivate the window
> manager. I'm using a button activated listbox to enter items into an
> application. The listbox is on a separate managerless Toplevel to give
> it a placard appearance.  When the listbox pops up, it has a local grab.
> Unfortanately, one can still drag around the main window underneath
> and thus inadvertantly hide the listbox.  I have managed to get around
> this problem by using withdraw/overrideredirect/deiconify on the
> main window when the listbox appears and again when the listbox
> disappears (to restart the window manager).  This is not a good fix
> since it results in irritating flashes as the main window is
> clobbered and redrawn.
> 
> Thus, my question - can one disable the window manager (or at least
> disable window movement) short of withdraw/overrideredirect/deiconify?
> It goes without saying that using grab_set_global is a bit too brutal.
> Thanks,
> Brian
> 

I've included some code that does a better job of it.  The new window
does actually use a global grab, just the same as a menu window.  The
behaviour matches how the same kind of window behaves in other more
complex widget sets.

==============================
#!/usr/bin/env python

from Tkinter import *
import string

class ScrolledMenu(Toplevel):
    def __init__(self, command=None):
        self.command = command
        Toplevel.__init__(self)
        self.withdraw()
        self.overrideredirect(1)
        self.listbox = Listbox(self)
        scroll = Scrollbar(self, orient='v', command=self.listbox.yview)
        self.listbox.configure(yscrollcommand=scroll.set)
        self.listbox.pack(fill='both', expand=1, side='left')
        scroll.pack(fill='y', side='right')
        self.bind('<ButtonPress-1>', self._press)
        self.bind('<ButtonPress>', self._close)
        self.bind('<KeyPress-Escape>', self._close)
        self.listbox.bind('<ButtonRelease-1>', self._select)
        self.up = 0

    def popup(self, onto):
        x = onto.winfo_rootx()
        y = onto.winfo_rooty() + onto.winfo_height()
        self.geometry('+%d+%d' % (x,y))
        self.deiconify()
        self.up = 1
        self.grab_set_global()

    def _close(self, event=None):
        self.grab_release()
        self.withdraw()
        self.up = 0

    def _select(self, event):
        self.update() # let the listbox change its selection
        self._close()
        index = self.listbox.curselection()
        value = self.listbox.get(index)[0]
        if self.command is not None:
            self.command(value)

    def _press(self, event):
        x,y = self.winfo_pointerxy()
        window = self.winfo_containing(x,y)
        if window is None or window.winfo_toplevel() != self:
            self._close()
            

class _Test(Tk):
    def __init__(self):
        Tk.__init__(self)
        self.sm = ScrolledMenu(self.result)
        apply(self.sm.listbox.insert, (0,)+tuple(string.lowercase))
        self.b= Button(self, text='Test', command=self.popup)
        self.b.pack()

    def result(self, value):
        print 'selected:', value

    def popup(self):
        self.sm.popup(self.b)

if __name__ == '__main__':
    mw = _Test()
    mw.mainloop()
==============================

--
Tim Evans



From curtis_c at lxe.com  Mon Nov 22 09:52:31 1999
From: curtis_c at lxe.com (Curtis, Craig M.)
Date: Mon, 22 Nov 1999 09:52:31 -0500
Subject: Python in DOS
Message-ID: <DB6390A9770ED211B2440000F807FC7B0103A127@nt_exlxe.lxe.com>

Commenting out those dreaded printfs certainly does work :-)

I guess I do not understand what PYTHONHOME should be set to.  I tried
setting it to the python-1.5.2 root and also to the lib under it, but
neither of these work.

Thanks,
Craig

> ----------
> From: 	Hans Nowak[SMTP:hnowak at cuci.nl]
> Sent: 	Friday, November 19, 1999 5:20 PM
> To: 	Curtis, Craig M.
> Cc: 	python-list at python.org
> Subject: 	RE: Python in DOS
> 
> 
> On 19 Nov 99, at 10:59, Curtis, Craig M. wrote:
> 
> > I tried setting PYTHONCASEOK to 0 and did the configure and make and
> still
> > have the same problem.  Any more suggestions?
> 
> Yeah, try setting it to 1. :o) 0 doesn't have much effect AFAIK. This 
> variable does not affect compilation though, it only affects the 
> interpreter when it runs.
> 
> If that doesn't help, do as I said... uncomment the code. I think 
> uncommenting a coupla fprintf statements in getpath.c will do. You'll 
> know 'em when you see 'em... they're these infamous error/warning 
> messages of course.
> 
> HTH,
> 
> --Hans Nowak (zephyrfalcon at hvision.nl)
> Homepage: http://fly.to/zephyrfalcon
> You call me a masterless man. You are wrong. I am my own master.
> 



From jblazi at netsurf.de  Mon Nov 15 16:03:39 1999
From: jblazi at netsurf.de (Janos Blazi)
Date: Mon, 15 Nov 1999 22:03:39 +0100
Subject: Beginning programming with Python
References: <sng0ychgwt.fsf@motorola.com> <382B3A53.5816B2@callware.com> <382B3D79.74CD3E0F@email.sps.mot.com> <80p9ds$ppo$1@newshost.accu.uu.nl>
Message-ID: <asteroidal-80psfo/INN-2.2.1/bat@broadway.news.is-europe.net>

I have tried to use Perl in my teaching. It has a very powerful but alas!
not very simple syntax! To start with, my pupils used to forget the dollar
signs and Perl does not like that. So this year I shall use Python. We shall
see. Time will tell, if Python is really better. Maybe my pupils will forget
the indenting this time? There is actually nothing else you can forget in
Python. I feel that it is the right language for teaching.

Janos Blazi
Germany


Martijn Faassen <m.faassen at vet.uu.nl> schrieb in im Newsbeitrag:
80p9ds$ppo$1 at newshost.accu.uu.nl...
> "Aaron Ginn (rp2885)" <rp2885 at email.sps.mot.com> wrote:
> > An honest opinion, please; is Python better suited for beginners than
Perl, as
> > I've heard?
>
> You're asking in the Python newsgroup, so here's my honest opinion:
>
> Yes.
>
> I was an experienced programmer (C, C++, Pascal, and some glancing
> familiarity with a host of other languages) when I learned Python.
> Before I looked at Python, I struggled a few weeks trying to teach
> myself Perl. Then I downloaded the Python tutorial, read it in about an
> hour, and I never looked back. I could program in Python from then on.
Python
> was much much easier for me to learn. I could reason about the right
> syntax for things, while with Perl I was constantly checking the
documentation.
>
> This was of course the example of an experienced programmer, but I think
> it'll be true for beginners too. The multiplicity of syntax of Perl to
> do the same thing, and the odd syntax to do certain things will confuse
> beginners, and will also teach the wrong idea to beginners; programming
> is not about syntax gimmicks. With perl you could be tempted to learn
> lots of snippets and tricks to solve a problem and how to assemble
> these to a working program. Lots of copy&paste programming, and lots of
> 'how do I do this in Perl' questions. With Python you can focus more on
> the underlying structure, the semantics, of the program.
>
> Then again, this is a biased opinion. There are arguments why Perl would
> suit the human mind better *because* it is so syntactically variable and
> complicated. I don't believe in these arguments myself.
>
> Python has a clean and powerful syntax. Perl has just a powerful syntax.
>
> Regards,
>
> Martijn
> --
> History of the 20th Century: WW1, WW2, WW3?
> No, WWW -- Could we be going in the right direction?





From thamelry at vub.ac.be  Wed Nov 24 03:48:56 1999
From: thamelry at vub.ac.be (Thomas Hamelryck)
Date: 24 Nov 1999 08:48:56 GMT
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk>
Message-ID: <81g8po$reo$1@mach.vub.ac.be>

Mark Carroll (markc at chiark.greenend.org.uk) wrote:
[snip]
: Basically, I'd be really interested to learn: what _don't_ people like
: about Python?
[snip]

Well, here are some of the things I don't like very much: 

Block delimitation by indenting. This makes python look like a porsche with
a caravan attached :-). As far as I know, most python users even seem to like
this for some reason.
I very much miss the unary arithmatic operations that C has (e.g. a+=1).
Lack of true garbage collection (which leads to problems with
circular references).
Lack of private attributes (although you can simulate private attributes by
picking a name with two underscores; python then does some name magling 
with it).

But of course, it's still a very nice language :-)

Cheers,

------------------------------------------------
Thomas Hamelryck//Free University Brussels (VUB)
Intitute for Molecular Biology// ULTR Department
Paardestraat 65//1640 Sint-Gensius-Rode//Belgium
------------------------------------------------



From tim_one at email.msn.com  Fri Nov  5 02:14:25 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Fri, 5 Nov 1999 02:14:25 -0500
Subject: ~ 1 = -2 ???
In-Reply-To: <m2so2lk08p.fsf@netmeg.net>
Message-ID: <000301bf275d$6468e120$1d2d153f@tim>

[Les Schaffer]
> ...
> so for my code, i will do
>
> def not4bit(x):
>       return ~x & 15
>
> >>> not4bit(1)
> 14

Les, if x is always in range(16), simply subtract it from 15.  That is,

def not4bit(x):
    assert 0 <= x < 16
    return 15 - x

It's twice as fast <wink>.

or-15-times-as-slow-it's-hard-to-tell-ly y'rs  - tim





From tim_one at email.msn.com  Thu Nov 25 05:38:58 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Thu, 25 Nov 1999 05:38:58 -0500
Subject: automatic "lotjes trekken"
Message-ID: <000001bf3731$486ea840$be2d153f@tim>

[Carel Fellinger]
> ...
> in the damp and cold days of late November and early December we
> think and ponder on a nice and teasing Sinterklaas gift for one-
> other. To prevent doublures, we draw lotjes (lots) with the name
> of the one to tease.  Ofcourse you are not allowed to draw one-self.
> And often some one would rather not have specific others; e.g. my
> eldest girls, being twins, rather not have each other, as the mere
> fact that one refuses to cooperate with the other is a sure sign
> that she drew her twin sister.  And part of the fun is that the
> drawing is secret and kept secret; though everybody tries to figure
> it out anyway.
>
> Being fed up with drawing and redrawing and redrawing and...
> I decided to use python.  Piece of cake using random.choice, sends
> mail to all and the lot.  Then I added these restrictions, still
> trivial. But elas running the little devil with to many restrictions
> took awile:) I decided to check before hand whether a drawing was
> possible at all by counting all possibilities. There must be something
> more clever than backtracking to get to this number, something with
> permutations and so, but my mind is to blurred to see it.

The mail<->news gateway is screwed up, so my apologies if you already got a
sufficient answer (although, as will become clear, I doubt it <wink>).

In combinatorics, this is the general problem of permutations with forbidden
positions:  you can view the lots as mapping the set of people into itself,
one-to-one and onto.  That is, it's a permutation.  Since a person is not
allowed to map to themself here, an upper bound is the number of
"derangements" (great word!) of N people.  A derangement is simply a
permutation where no element maps to itself.  Asymptotically, there are N!/e
derangements of N things (N! is N factorial = N * (N-1) * ... * 1, and e is
2.71828...).  This is an excellent approximation even for small N; e.g., for
N=5, 5!/e = 120/e ~= 44.15, while the exact number of derangements of 5
things is 44.

That the number e pops out of a very uniform special case should suggest
that the general theory isn't likely to be dirt simple.  And, in a great
surprise, it turns out that it actually isn't dirt simple <wink>.  Something
you could ponder the whole winter!

To get you started, here's a simple (& std) way to think about it:  picture
an NxN chessboard.  A permutation corresponds naturally to placing N rooks
on the chessboard in such a way that no rook can capture another.  For
example, label the rows of the chessboard with the names of the people, and
the columns likewise, and say that a rook in rank i and file j (row i and
column j) means that person i has drawn person j.  No two rooks can be in
the same row (a person can't get gifts for two people) or in the same column
(a person can't get gifts from two people).

A permutation with forbidden positions is then placing N non-capturing rooks
on a *damaged* chessboard, where large irregular lumps of hardened smelly
cheese are firmly stuck to forbidden squares, thus preventing a rook from
standing on them.  For example, the derangement requirement is modeled by
placing cheese all along the main diagonal.

Now you *can* systematically count the number of ways to do this without
backtracking, but accounting for the 2-dimensional geometry of the forbidden
positions is essential.  This makes it tricky.  Here's how:

Stare at your damaged chessboard.  Pick any square without a lump of cheese
on it (if there aren't any, you're done:  it's impossible).  You have two
choices:  you can either put a rook there, or not.  This splits all possible
solutions into two mutually exclusive kinds, so the total number of
solutions is the sum of the number of solutions for each case:

1. You do put a rook there.  Then physically remove that square's rank and
file from the chessboard, and count the number of ways to put N-1
non-capturing rooks on the resulting N-1 x N-1 chessboard.

2. You do not put a rook there.  Then firmly press a large irregular lump of
hardened smelly cheese on to the square, and count the number of ways to
place N non-capturing rooks on this cheesier N x N chessboard.

3. Add those two results.

This reduces the problem to two subproblems of the same kind, but each
subproblem either has a smaller board to worry about or a board with more
cheese.  So proceed recursively, and every path eventually leads to a 1x1
board and/or a board completely filled with cheese.  Those are trivial.

Here's the depressing part:  the end cases are *too* trivial.  The tails of
the recursion contribute only 0 or 1 each to the total count.  So there are
a *lot* of function calls to get those to add up to something that can grow
as large as N!/e (if you've got no restriction other than derangement).

This is where "accounting for the 2-dimensional geometry" comes into play:
at each step, you want to pick the free square with care, striving to
produce subproblems where the cheese-free squares can be split into two
smaller chessboards with no rows or columns in common.  The number of
solutions for a chessboard of that special form is the *product* of the
number of solutions for the two smaller chessboards (they have no rows or
columns in common, so what you do to one of them can have no effect on the
other), and that lets you get to the final answer enormously quicker.

The next twist in the tale is too involved to explain here:  it turns out
that it's possible to deduce the result from the numbers of ways to place 0,
1, 2, ... and N non-capturing rooks on the "cheese-inverted" chessboard
obtained by putting cheese on all the empty squares and removing the cheese
from the cheesy squares.  In most applications, there are many more empty
squares than cheesy squares, so inverting their roles can slash the amount
of computation.

If you enjoy this sort of thing, the theory is elegant and makes for a
delightfully delicate series of programming challenges (start with a Cheese
class <wink>).  I did it in Fortran once, many years ago -- and that I don't
have to anymore is the true meaning of Thanksgiving <gobble>.

sinterturkeyly y'rs  - tim





From sessile at in-gen.net  Tue Nov  9 13:16:39 1999
From: sessile at in-gen.net (sessile at in-gen.net)
Date: Tue, 09 Nov 1999 13:16:39 -0500
Subject: Can't use len() on returned PyList object?
In-Reply-To: <19991109143023.A1001@Ridcully.home>
References: <3.0.6.32.19991108211217.007cb210@mail.in-gen.net>
 <3.0.6.32.19991108211217.007cb210@mail.in-gen.net>
Message-ID: <3.0.6.32.19991109131639.007d89f0@mail.in-gen.net>

My thanks to those who responded.  You were correct.  I had
assigned 'len' as a variable elsewhere in the program.  I quite
foolishly did this by too closely emulating the C source that
my Python program is going to replace.  Doh!

Thanks again.

At 02:30 PM 11/9/1999 +1100, you wrote:
>On Mon, Nov 08, 1999 at 09:12:17PM -0500, sessile at in-gen.net wrote:
>> Why can't I use 'len()' to find the length of a list
>> returned from my new extension module?  The error is
>> shown as:
>> 
>>     TypeError: call of non-function (type int)
>
>You have probably used 'len' as an integer variable at some point
>previously in the program. This would have overwritten the builtin
>declaration of len as a function.




From tbryan at python.net  Mon Nov  8 21:47:59 1999
From: tbryan at python.net (Thomas A. Bryan)
Date: Mon, 08 Nov 1999 21:47:59 -0500
Subject: python: something completely different
References: <3820C0B8.62B2@iowegian.com> <38249ED2.BB7EA1DE@znet.com> <8028jv$7cf$1@hub.org> <38259F18.437@iowegian.com>
Message-ID: <38278B5F.1AFF953C@python.net>

Grant Griffin wrote:

> I don't know if a Matlab-to-Python translator is currently available,
> but if not, I don't think it would be all that hard to write one (in
> Python, of course).

I'm sure how useful such a thing would be.  Most people I know who 
seriously use Matlab use at least one of its toolboxes.  They also 
tend to output plots/graphs when they're done with their work, and 
Matlab actually makes some very nice plots.  I haven't looked at the
state of Python plotting/graphing in about a year, but when I last 
looked, I judged that the effort to produce nice plots with Python 
would be considerable.  

That said, from what I remember of Matlab syntax, it probably would 
be a fairly straight-forward project to implement a translator for 
the basic vector/array/structure/scripting syntax (assuming that 
we're translating to a Python with Numeric Python compiled).  It 
might be a fun project, but it probably wouldn't move too many 
people away from Matlab until they could get colored, labeled, 
scaled plots from Python.  Most people would also have to be willing 
to code the few crucial routines that they've been using from a 
Matlab toolbox.  Of course, if the translator existed, maybe 
more people would try to move to Python.  That could spur a 
lot of growth in toolboxes for Numeric Python! :)

---Tom



From ivanlan at callware.com  Mon Nov 29 19:04:56 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Mon, 29 Nov 1999 17:04:56 -0700
Subject: [Tutor] TKinter
References: <003301bf3ab5$13138080$0c5bdfc8@the-user> <38431229.7B60A603@worldnet.att.net>
Message-ID: <384314A8.6929482C@callware.com>

Hi All--

Ralph Alberti wrote:
> 
> Have you seen this:
> 
> http://starship.python.net/crew/fredrik/tkintro/Introduction.htm
> 
>   --Ralph
> 
> Alexandre Passos wrote:
> >
> > Has someone got a link to some __downloadable__ tutorial of TKinter?
> >

And also this:

	http://www.pythonware.com/library.htm

Which has a link to the PDF version of the introduction.

There're lots of parts missing, 'cause /F is so busy, but it's still
about the best place to start. ...

Take care,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours
----------------------------------------------



From wware-nospam at world.std.com  Tue Nov 30 09:12:49 1999
From: wware-nospam at world.std.com (Will Ware)
Date: Tue, 30 Nov 1999 14:12:49 GMT
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <81bmmi$279$1@nnrp1.deja.com> <3841EE77.6B5D6AC9@wjk.mv.com> <FLxz96.IEI@world.std.com> <38435C43.B238EC15@mindspring.com>
Message-ID: <FM0KtD.2ps@world.std.com>

Will Ware wrote:
> I find myself grumbling about having to type "x = x + 1". The really
> clean thing to do, given that integers are objects, would be to define
> increment and decrement methods, so you'd type something like "i.incr()".

Jesse D. Sightler (jsight at mindspring.com) wrote:
> Wouldn't this be impossible, since Integers are also immutable?  So, eg,
> i.incr() could only return an incremented version of i, but not actually
> increment i?

Right, this is what Gordon was pointing out. A "++" method could not
magically reach out of its object, find the "i" variable, and bind it
to the incremented value. Something like this could be accomplished
by tinkering with Python's parser, but that wouldn't be a great idea.
A lot of work for something that would never find wide acceptance.
-- 
 - - - - - - - - - - - - - - - - - - - - - - - -
Resistance is futile. Capacitance is efficacious.
Will Ware	email:    wware @ world.std.com



From exeq at my-deja.com  Wed Nov  3 17:06:39 1999
From: exeq at my-deja.com (exeq at my-deja.com)
Date: Wed, 03 Nov 1999 22:06:39 GMT
Subject: Tkinter and Solaris 7 Sparc [Q]
Message-ID: <7vqblc$ham$1@nnrp1.deja.com>

Hi!

I am trying to compile Python 1.5.2 with tkinter on Solaris 7 (Sparc). I
have installed TCL/TK 8.0.5, GCC 2.8.1 / Sun Workshop 5 (I tryied with
both compilers).

I just get an error saying that libX11.so.6.1 is needed. Is there a way
to build tkinter without X11? Or I just need to download and install
X11?

Thanks in advance.

/B


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



From mcalla at home.com  Tue Nov  9 20:49:06 1999
From: mcalla at home.com (Mike Callahan)
Date: Wed, 10 Nov 1999 01:49:06 GMT
Subject: Using GNU C complier as the "default" windows complier
Message-ID: <ma4W3.7794$Ur4.223420@news.rdc1.tn.home.com>

This a question from a newbie but I was thinking...why doesn't the default
Python code for windows use the GNU C complier for windows instead of the MS
complier. Would it be possible to distribute the GNU package along with the
Python package and thus make it easier to compile, extend, and embed Python,
use freeze, etc? I know it would take some work, but it seems it would make
windows Python a lot more like unix Python. Just a thought...





From thantos at chancel.org  Tue Nov 16 18:05:13 1999
From: thantos at chancel.org (Alexander Williams)
Date: Tue, 16 Nov 1999 23:05:13 GMT
Subject: Beginning programming with Python
References: <sng0ychgwt.fsf@motorola.com> <382B3A53.5816B2@callware.com> <382B3D79.74CD3E0F@email.sps.mot.com> <80p9ds$ppo$1@newshost.accu.uu.nl> <asteroidal-80psfo/INN-2.2.1/bat@broadway.news.is-europe.net> <38312D55.A36EB56D@iqsoft.hu> <antigorite-80rol2/INN-2.2.1/attentive@broadway.news.is-europe.net>  <383195C1.826B3342@iqsoft.hu>
Message-ID: <slrn833ot6.83i.thantos@chancel.org>

On Tue, 16 Nov 1999 18:34:57 +0100, Nemeth Miklos <nemeth at iqsoft.hu>
wrote:

>engineer. I like Python but I'd like to be convinced that Python is not only
>for scientific programming, but also for business programming. I'd like to know
>the general opinion of the "official" Python community on this (I think)

The proof is in the pudding, as it were; there are several /very/
large programming projects in the business community (including my
own) which are Python Powered; I've certainly found maintainance
/easier/ in Python than in strongly-typed languages; a type-error will
typically be found in testing (which is far easier to do bit by bit as
you go with the interpreter than the usual compile-run-edit cycle)
rather than when the program is deployed.


-- 
Alexander Williams (thantos at gw.total-web.net)
"In the end ... Oblivion Always Wins."



From m.faassen at vet.uu.nl  Sun Nov 14 06:14:45 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 14 Nov 1999 11:14:45 GMT
Subject: Python for children.....
References: <9911051740130D.02530@william.jinsa.org> <7GOU3.272$Ae3.194028@news-dal.corridex.com> <801p16$10s$1@vvs.superst.iae.nl> <199911080328.WAA24200@acer.spvi.com> <199911080333.WAA24231@acer.spvi.com> <809tlb$1eg$1@vvs.superst.iae.nl>
Message-ID: <80m5j5$mvm$2@newshost.accu.uu.nl>

Carel Fellinger <cfelling at iae.nl> wrote:
> Steve Spicklemire <steve at acer.spvi.com> wrote:

>> Forgot to mention:

>> PyBots: http://opensource.spvi.com

> looks interesting, but elas, it doesn't run on linux
> and i've hardly any room on my left-over windows partition.
> so I've to wait for an other port or Sinterklaas.

Ah, yeah, Sinterklaas, Sinterklaas. I explained all about him last year
when the Goedheilig Man was around. Interested people should check DejaNews. :)

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?



From cgw at fnal.gov  Tue Nov 16 07:14:32 1999
From: cgw at fnal.gov (Charles G Waldman)
Date: Tue, 16 Nov 1999 13:14:32 +0100 (MET)
Subject: Alright, I give up...chalk this one up to experience...
In-Reply-To: <383145c0.4445469@212.225.36.99>
References: <383145c0.4445469@212.225.36.99>
Message-ID: <14385.19112.173132.954102@gargle.gargle.HOWL>

Endgamer writes:

> I can't see why, when I run this as a script from Windows 95, it
> doesn't return the current date setting on my machine.
>
> # Date setting program (chunk 1)
> import os
> 
> olddate=os.popen("date",'r')
> olddate=readdate
> print olddate
> 


What is "readdate"?  I'd expect this snippet of code to give a
NameError since readdate isn't defined anywhere.  I'm not quite sure
what your intention is here, with the line "olddate=readdate"; what's
the point of calling the popen function and assigning the returned
value to olddate if on the very next line you're going to throw this
value away (by reassigning to olddate), before doing anything with it?

I think the info that you need to know is this: os.popen returns a
file-like object from which you can read the output of the executed
command.  You use it like this:

import os
pipe = os.popen("date",'r')
olddate = pipe.read()
exit_status = pipe.close()
print olddate




From nemeth at iqsoft.hu  Thu Nov  4 14:08:32 1999
From: nemeth at iqsoft.hu (Németh Miklós)
Date: Thu, 04 Nov 1999 20:08:32 +0100
Subject: Extending Python dynamically
References: <3F8BA119B48AD11197470008C7A493CB98B636@HQOBGEX1>
Message-ID: <3821D9B0.56C01339@iqsoft.hu>

ScherBi at BAM.com wrote:

> > I have downloaded DCOracle extension. I have RH6.0, and (at
> > least until now) I have not installed the source distribution of Python. I
> have
> > installed only the binary RPMs. I unpacked DCOracle in a directory and
> > managed to build the so files (Buffer.so and oci_.so).
> > How to install then this DCOracle package? I have been trying
> > to figure out but no success.
>
> Create a directory somewhere in your pythonpath (import sys; print
> sys.path), call it DCOracle.
> In this directory put the two .so files, and all the .py files that came in
> the DCOracle package.
>
> Then you may import DCOracle.
>
> Let me know if you need more help.
>
> Bill Scherer

Thank you for your help. The method you suggested works perfectly!
However, I do not understand completely how the Python interpreter finds out
that it has to load an so file. Normally -- I believed -- when Python sees an
"import xxx" statement it looks for a xxx.py, xxx.pyc or xxx.pyo file in the
$PYTHONPATH directories or an xxx subdirectory having an __init__.py file, but
not an xxx.so file. This means that if I add the current directory to the
$PYTHONPATH (such as export PYTHONPATH='.' Python will find any arbitrary so
file in the $PWD?
Anyway, I am very happy that I do not have to recompile Python every time I add
a new extension library (ie. an so file) :-).
Thanks again,
NM





From guido at cnri.reston.va.us  Thu Nov  4 22:36:44 1999
From: guido at cnri.reston.va.us (Guido van Rossum)
Date: 04 Nov 1999 22:36:44 -0500
Subject: __str__ vs. __repr__
References: <000401bf25b9$88ea57e0$2e2d153f@tim> <199911040157.UAA11908@eric.cnri.reston.va.us> <m31za6339e.fsf@greebo.nodomain.de>
Message-ID: <5logd98v8z.fsf@eric.cnri.reston.va.us>

Bernhard Herzog <herzog at online.de> writes:

> How about a hook function a la __import__? I.e. have a function
> __format__ [1] in __builtin__ that is bound to repr by default. Whenever
> Python prints the result of an expression in interactive mode, it calls
> __format__ with the result as parameter and expects it to return a
> string ready for printing.

Brilliant!  The default function could do all the current built-in
magic -- print nothing if it's None, assign it to builtin _, and so
on.

I wonder if it would have to have a __magic__ name?  It could be
called display and it could be documented and usable in
non-interactive programs as well.  Or am I getting carried away?
(Possibly the habit if assigning to _ and suppressing None would make
the default display() function a bit clumsy to use.)

--Guido van Rossum (home page: http://www.python.org/~guido/)



From tim_one at email.msn.com  Wed Nov 10 00:30:13 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Wed, 10 Nov 1999 00:30:13 -0500
Subject: IDLE and rlcompleter [Was: Re: making the Python case to my dept.]
In-Reply-To: <m366zcoagc.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <000201bf2b3c$a977dc20$d32d153f@tim>

[Michael Hudson]
> ...
> While we're on the subject of idle, is there any easy (preferably one
> or two keystrokes) that way to import/reload the module being edited
> into the Python shell and bring said shell to the front for instant
> interactive play? That'd be ?bercool.

Already implemented, in the current CVS IDLE.

ubergooberly y'rs  - tim





From ajung at sz-sb.de  Thu Nov 11 10:38:38 1999
From: ajung at sz-sb.de (Andreas Jung)
Date: Thu, 11 Nov 1999 16:38:38 +0100
Subject: .html -> .ps ?
In-Reply-To: <19991111080506.A1007@better.net>; from parkw@better.net on Thu, Nov 11, 1999 at 08:05:06AM -0500
References: <19991111080506.A1007@better.net>
Message-ID: <19991111163838.A943@sz-sb.de>

On Thu, Nov 11, 1999 at 08:05:06AM -0500, William Park wrote:
> Hi,
> 
> I need to convert HTML file to Postscript, so that I can send fax it.
> 
> 1.  Does anyone have experience with 'html2ps', 'html2tex', and
>     'html2latex' programs?  Or, is there any other program that will do
>     the conversion in Python or otherwise?  
> 
> 2.  Is there a command-line options to tell Netscape to print to a file?
>     This would be ideal, since one can already do this manually with
>     mouse.

Take a look at w3m (http://ei5nazha.yz.yamagata-u.ac.jp/~aito/w3m/eng/ ).
w3m is a textmode browser like lynx however w3m is able to render tables.
With the -dump option of w3m you can dump the ascii outpout to stdout - 
so you're able to convert html to ascii and then to postscript. There
is also a html2ps converter (in Perl written) available however
the latest version is outdated and has never readched a stable state.

Cheers,
Andreas 



From olivier at scripps.edu  Thu Nov 11 10:57:34 1999
From: olivier at scripps.edu (Olivier Roche)
Date: Thu, 11 Nov 1999 07:57:34 -0800
Subject: How do I compile NumPy on Linux?
References: <382A8094.3354DE70@mdmlab.mi.infm.it> <19991111064736.A528@better.net>
Message-ID: <382AE76D.A448CC4B@scripps.edu>

William Park wrote:

> On Thu, Nov 11, 1999 at 09:38:45AM +0100, Sandro Ferrari wrote:
> > I downloaded the LLNLDistribution.tgz file and followed the installation
> > instruction:
> > cd Numerical
> > ./makethis.py
> > but I get the message:
> > make: Makefile.pre.in: No such file or directory
> > make: *** No rule to make target `Makefile.pre.in'.  Stop.
> > make: *** No targets.  Stop.
> > In the documentation isn't mentioned anything about it.
> > Any help?
> > thanks
> > Sandro
>

You must just install the "python-devel" rpm to get the Makefile.pre.in.
You should have with rpm -qa | grep python
python-1.5.1-10
pythonlib-1.23-1
python-devel-1.5.1-10

Olivier

--
___________________________________________________________

Olivier Roche, PhD
Department of Molecular Biology, TPC6
The Scripps Research Institute            Tel (858) 784 9782
10550 North Torrey Pines Road             Fax (858) 784 8688
La Jolla,  California  92037     E-mail: olivier at scripps.edu
http://www.multimania.org/09/oroche/
____________________________________________________________






From S.I.Reynolds at cs.bham.ac.uk  Wed Nov 17 07:02:54 1999
From: S.I.Reynolds at cs.bham.ac.uk (Stuart Reynolds)
Date: Wed, 17 Nov 1999 12:02:54 +0000
Subject: Wild idea - a type finder
Message-ID: <3832996E.1517@cs.bham.ac.uk>

Does anyone know of any tools which let you do things like:

We have some exisiting function:

  def f(a, b):
      return a<b

And we want to reverse engineer its type from its implementation:

   findtype( f )

e.g.
- function takes 2 args
- a and b need to be comparable
- returns a boolean (actually an int in Python)

So the type of f is:

 (comparable)(comparable)-->(int)

Or in another example you might get:

  >>> findtype( add )
  (number)(number)-->(number)

  >>> findtype( map )
  (()())(seq)(seq)-->(seq)
  
( ()() ) denotes a function which takes two arguments of unknown type.


I've only ever seen this kind of thing in pure functional programming
languages (gopher to be specific).

The reason I ask is that I think, in general, its really useful to be
able to tell what minimum interface the arguments you give to a method
needs to support (you can stick it in your documenation for one thing).
It would also be pretty useful to uncover the kind of nasty bugs which
crop up in untyped langauges whereby the functions or methods you've
written have stronger requirements on the interfaces of their arguments
than you realised (espcially in the case where your code calls lots of
other functions who's types you're not too sure about).


Stuart



From zippy at cs.brandeis.edu  Wed Nov 24 01:40:21 1999
From: zippy at cs.brandeis.edu (Patrick Tufts)
Date: Tue, 23 Nov 1999 22:40:21 -0800
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <38397696@news.xtml.co.uk> <221119992150405215%zippy@cs.brandeis.edu>
Message-ID: <231119992240219368%zippy@cs.brandeis.edu>

In article <221119992150405215%zippy at cs.brandeis.edu>, Patrick Tufts
<zippy at cs.brandeis.edu> wrote:

> Here's an example:
> 
> (defun foo (x) (mod x 4))    ; return x mod 4, always an integer
> 
> (defun bar (x) (^ 2 x))      ; return 2^x, always an integer
> 
> (defun baz (a b) (+ (foo a) (bar b)))  ; foo(a) + bar(b)

A reader has pointed out that this is a bogus example, because foo and
bar are not guaranteed to return ints (among other things).

I could give an example like 

(defun foo (x) 4)
(defun bar (x) 3)
(defun baz (a b) (+ (foo a) (bar b)))

but that's so simple that it's not very interesting, and my mind is too
fuzzy now to come up with an interesting example.

If anyone else is up to the challenge of correctly saying what I should
have, please do.

--Pat



From fredrik at pythonware.com  Thu Nov 11 03:17:15 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 11 Nov 1999 09:17:15 +0100
Subject: How to perform syntax check in Python
References: <382A2384.CC37FBA8@hp.com>
Message-ID: <00bf01bf2c1d$2a927be0$f29b12c2@secret.pythonware.com>

Meera <meera_bavadekar at hp.com> wrote:
> I am looking for  a utility to perform syntax validation in python.
> Consider following code -
> 
> try:
>   a = 3
>   print a
>   #Do some more processing
> except:
>   print b
> 
> The fact that variable b is not defined is known only when we get
> exception.
> We will get 'NameError exception' _not_  the actual exception we are
> expecting.
> 
> When we have multi-developer and multi-platform environment - such
> feature is crucial.  This is one area which I would like python to improve
> over conventional language.

having been there and done that, I'd say a good testing
harness is a better way to address that -- helps you find
both code glitches and real bugs (!).

but if you really prefer linting, check out chordate's pylint 
utility, available from:

    http://www.chordate.com/kwParsing/index.html

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
(news: just posted an errata for the 'first printing') -->




From dway at mat.net  Wed Nov  3 19:43:06 1999
From: dway at mat.net (Doug Way)
Date: Thu, 04 Nov 1999 00:43:06 GMT
Subject: Python and Smalltalk (Was: Re: Comparison between Python and "Ruby")
References: <1270663667-24391282@hypernet.com> <381ED7A6.D7D7FB13@compaq.com> <877lk1wiij.fsf@ev.netlab.co.jp> <19991102154543.A4469@optiplex.palga.uucp> <381F9EB4.63FA033F@python.net> <slrn81ve6m.l1c.neelk@brick.cswv.com>
Message-ID: <7vqkqq$oa7$1@nnrp1.deja.com>

In article <slrn81ve6m.l1c.neelk at brick.cswv.com>,
  neelk at alum.mit.edu wrote:
>
> Python (and Ruby, for that matter) remind me more of Smalltalk than
> anything else -- ie, they have extensive libraries, a community with a
> love of elegance but not of purity[*], and a greater focus on
practical
> applications than on academic research.
>
> The differences are that there's less of a closed-world attitude in
> Python than in the ST world, though -- Python can be used for sysadmin
> work where ST (even Squeak) wouldn't be a natural fit. But on the flip
> side Smalltalk has always had a stronger focus on UI and dev
> environments.  This is changing, though -- the ST culture is beginning
> to value the ability to deal with the world outside the system image,
> and Guido's latest enthusiasm is "computer programming for everybody"
> (which is a very ST-ish sort of idea).
>
> Neel

Smalltalk is sort of on the "other side" of Python from Perl in a lot of
ways.  Smalltalk's syntax is considerably pared down from even Python's
not-too-complex syntax.  In other words, if you hate Perl, you might
really like Smalltalk. :-)  (Randal Schwartz might disagree with me
here, 'cause I know he's on the Squeak list...)

I agree with Neel about the differences that exist.  Python does
interact with the outside world better, but Smalltalk's dev tools really
are great, and are easy to tweak to suit your own needs, since
everything's written in Smalltalk and the source is right there.

>From a historical point of view, Smalltalk is a little like Lisp, in
that it's a little outside of the mainstream, but it has enough new
ideas in it that will give it a very long lifespan.  Smalltalk will
still be relevant 20 years from now, just as Lisp-based languages are
still reasonably relevant today, after 40 years.  I suppose "having a
lot of new ideas" and "being a little outside of the mainstream" go hand
in hand.  (Smalltalk isn't *that* obscure, though... there are more
Smalltalk programming jobs available than, say, Python or Lisp jobs.)

Contrast this with a mainstream language like Java, which will have a
huge spike in popularity for a few years, but since it's mostly just a
cleaned-up version of another language (C++), it won't be cutting-edge
for very long.  Which isn't to say that it will dissappear, just that it
will be easier to replace with something better as the new "King of the
Mainstream Programming Languages" in a few years.

In any case, Smalltalk is definitely worth exploring if you're looking
at programming languages similar to Python.  See http://squeak.org,
http://www.goodstart.com/dialects.html, etc.

- Doug Way
  EAI/Transom Technologies, Ann Arbor, MI
  http://www.transom.com
  dway at mat.net, @eai.com


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



From sekter at mail.matav.hu  Mon Nov 15 06:46:03 1999
From: sekter at mail.matav.hu (Arpad Kiss)
Date: Mon, 15 Nov 1999 12:46:03 +0100
Subject: PIL 1.0 on win32?
Message-ID: <009b01bf2f5e$fef3b900$029f38c3@geometria.hu>

Hi,
I have a PIL based Zope product I have been developing on a Linux box. I
just wanted to test it on a Windows NT(sp5), but it doesn't work.:-(
At first it didn't recognize my fonts, because the readline method on win32
returns 10+13 at the end of the lines and in ImageFont.py only \n is
checked.
After I  solved it I got an error message(memory error in line 151 of
ImageDraw.py) when I wanted to draw some text.
When I commented out the text drawing code in my program, then I got back a
picture, but it is mirrored!
Of course on linux box my program works fine.
I just want know who else use PIL 1.0 on win32 and what experiences they
have. Should this text drawing function work? Is it my fault?
Best regards,
Arpad Kiss






From skaller at maxtal.com.au  Thu Nov 25 17:48:20 1999
From: skaller at maxtal.com.au (skaller)
Date: Fri, 26 Nov 1999 09:48:20 +1100
Subject: Python on top of Oberon instead of C?
References: <8192fu$91h$1@nnrp1.deja.com> <slrn83guke.o5d.wtanksle@hawking.armored.net>
Message-ID: <383DBCB4.91CB8B5F@maxtal.com.au>

William Tanksley wrote:
> 
> On Sun, 21 Nov 1999 15:18:24 GMT, srenner at lycosmail.com wrote:
> >Anyone think this is an interesting idea?
> 
> Not as such -- the work of translating Python's existing code to Oberon
> doesn't seem to be repaid by the benefits of Oberon.
> 
> OTOH, I _would_ like to see a Python compiler which output Juice format
> rather than bytecodes.  That would be very nice indeed.  My understanding
> is that Viper is very close to this already.

Not yet. In a previous article, Hal wrote:

>Having already responded on the Native Oberon list, I must add that I'm more
>familiar with Oberon than Python.  But I'm back-and-forth between these two
>languages.  I'm guessing that the OPython interpreter that produces the byte
>code (*.pyc) would be the hardest part.

I don't think this is actually correct. My experience building Viper is
that
the 'hardest' part is implementing the libraries. Not that it is 'hard'
per se, just quite laborious, and requiring continual fiddles to the 
overall architecture, while juggling compromises between CPython
implementation details, and the needs of real CPython programs.

I consider the following modules basic: _builtins_, string, exception,
types,
marshal, pickle, sys, os, posix, re, array, traceback, time (and a few
more ..)
Interestingly, string, exception, and types work as written in pure
python, 
without modification (but 'string' is way too slow). 
In addition, Viper will support a full scale, _builtin_ GUI, rational
numbers, and probably matrices.

I'd guess and Oberon version of Python would face the same issues:
I don't think that the 'python language' is much use without the
standard
libraries people need to do actual work.

-- 
John Skaller, mailto:skaller at maxtal.com.au
10/1 Toxteth Rd Glebe NSW 2037 Australia
homepage: http://www.maxtal.com.au/~skaller
voice: 61-2-9660-0850



From pinard at iro.umontreal.ca  Mon Nov 29 17:28:16 1999
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 29 Nov 1999 17:28:16 -0500
Subject: Proposal: Python Info Collective
In-Reply-To: m.faassen@vet.uu.nl's message of "18 Nov 1999 10:20:40 GMT"
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <JlFX3.3659$s87.67930@dfiatx1-snr1.gtei.net> <80oqn9$ocq$2@newshost.accu.uu.nl> <oLUX3.286$yA4.4821@dfiatx1-snr1.gtei.net> <80p9su$ppo$2@newshost.accu.uu.nl> <80piru$13rl$1@hub.org> <80re5d$55a$1@newshost.accu.uu.nl> <fjPY3.770$An3.14793@newsfeed.avtel.net> <810jto$lo0$1@newshost.accu.uu.nl>
Message-ID: <oq903hdjb3.fsf@titan.progiciels-bpi.ca>

m.faassen at vet.uu.nl (Martijn Faassen) writes:

> No, no, a more extensive deja search indicated to me that this person
> has posted under other names to the group previously.

Sh*t, I'm unveiled!

God knows I did try my very best at using different writing styles,
consistently, for each of the various names I use when I wrote to the list,
so far.  And especially, when I write as Tim Peters.  I may confess it now,
this invention of mine has been fairly difficult to sustain, all along :-)

P.S. - Odo, to one Bajoran: "Your noses are giving me some trouble."

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





From fredrik at pythonware.com  Mon Nov 15 07:04:24 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 15 Nov 1999 13:04:24 +0100
Subject: PIL 1.0 on win32?
References: <009b01bf2f5e$fef3b900$029f38c3@geometria.hu>
Message-ID: <01ee01bf2f61$96097a80$f29b12c2@secret.pythonware.com>

Arpad Kiss wrote:
> I have a PIL based Zope product I have been developing on a Linux box. I
> just wanted to test it on a Windows NT(sp5), but it doesn't work.:-(
> At first it didn't recognize my fonts, because the readline method on win32
> returns 10+13 at the end of the lines and in ImageFont.py only \n is
> checked.

PIL opens the font files in binary mode, so this
shouldn't be a problem.  you probably messed
something up when you transferred the font
files to the PC...

> After I  solved it I got an error message(memory error in line 151 of
> ImageDraw.py) when I wanted to draw some text.
> When I commented out the text drawing code in my program, then I got back a
> picture, but it is mirrored!

this sounds really weird -- but at least the
memory error might be due to bogus font data...

what win32 build do you use?  can you post some
code?

> Of course on linux box my program works fine.
> I just want know who else use PIL 1.0 on win32 and what experiences they
> have. Should this text drawing function work? Is it my fault?

fwiw, PIL is being developed on win32...

</F>




From wjk at wjk.mv.com  Sat Nov 27 12:54:27 1999
From: wjk at wjk.mv.com (William J. King)
Date: Sat, 27 Nov 1999 12:54:27 -0500
Subject: SMTP
References: <19991126095821.87186.qmail@hotmail.com>
Message-ID: <38401AD2.31396C84@wjk.mv.com>

Was just looking at an interesting program called Rebol which does some
interesting things with mail/ftp etc. http://www.rebol.com - handles
various protocols etc. A scripting communications/messaging language..
hope I got that correct.. Sassenfrath of Amiga fame is source.

This might be yet another simple solution for  SMTP area.




From mark.hamstra at bentley.com  Sat Nov  6 14:12:50 1999
From: mark.hamstra at bentley.com (Mark Hamstra)
Date: 06 Nov 1999 14:12:50 -0500
Subject: JPython and NetBeans and Java envy (Was Re: Python and Smalltalk and Java envy)
References: <1270663667-24391282@hypernet.com> <7vqkqq$oa7$1@nnrp1.deja.com> <382484F1.F8B8D712@home.com> <99110612201506.02970@quadra.teleo.net>
Message-ID: <xzaeorigct.fsf@sullivan.bentley.com>

Patrick Phalen <python-list at teleo.net> writes:

> [Wostenberg, on Sat, 06 Nov 1999]:
> 
> :: Intersting observation. Perhaps it comes down to those of us who are
> :: comfortable with niche languages (Python, Smalltalk, Lisp) and those who
> :: suffer Java-envy . Do Python people, Ada advocates, or little Lispers wring
> :: there hands about the fact "we are not the most used language?" Why should
> :: Smalltalkers?
> 
> 
> Speaking of Java envy, Barry Warsaw pointed out the other day on the
> JPython list that www.netbeans.com is running a survey to pick the best
> scripting language to be integrated with NetBeans. (You'll find the form
> in their left navbar.)
> 
> I won't reveal the *overwhelming* winner, thus far. Find out for
> yourself.  ;)


Just curious, why not Bean Scripting Framework, since it allows for JPython
and any other scripting language that has been hooked into the BSF?
 
-- 
Mark Hamstra
Bentley Systems, Inc.



From x at x.x  Wed Nov 24 17:45:47 1999
From: x at x.x (Eric Jacobs)
Date: Wed, 24 Nov 1999 17:45:47 -0500
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <867lj85g4c.fsf@g.local> <al8yabo9fm3.fsf@sirppi.helsinki.fi> <00bf01bf3685$0c345610$f29b12c2@secret.pythonware.com>
Message-ID: <383C6A9B.4049CB12@x.x>

Fredrik Lundh wrote:
>
> note that python already has a special syntax
> for this purpose:
> 
> >>> a = range(20)
> >>> print a[1:20:3]
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
> TypeError: sequence index must be integer
> 
> (note that you got a type error, not a syntax
> error.  numpy arrays fully support this syntax).
> 
> in other words, you don't have to be guido to
> realize that something like:
> 
>     for i in [1:20:3]:
>         ...
> 
> could be the right way to do it in Python.  in
> fact, it's already on the list for 2.0:

I don't know if I like that syntax. Taking a slice
of what? If I had to replace range, I'd go with:

   for i in Int[1:20:3]:
      ...

Here Int is an object that represents a list of all
integers. We take a slice of it... and get exactly
what we expect.

Either way, the semantics of lists with an infinite
number of elements or negative indices would have
to be cleared up somewhat. That would be a good
thing.



From poinot at onera.fr  Thu Nov 25 03:28:38 1999
From: poinot at onera.fr (Marc POINOT)
Date: Thu, 25 Nov 1999 09:28:38 +0100
Subject: Numeric module under IRIX 6.5
References: <383C1D37.41863D9B@calibre-dd.com>
Message-ID: <383CF336.1B6ED368@onera.fr>

Stephen J King wrote:
> 
> I've downloaded LLNL Numeric module and compiled it for IRIX 6.5 (using
> -n32 flag), but when importing the module to Python, it dumps core |=(
> 
> Has anyone encountered/solved this problem?
> 
> Thanks
> 
> --  Stephen J King    Technology Manager   Calibre Digital Pictures
> --  king at calibre-dd.com               http://www.calibredigital.com
> 
> --
> http://www.python.org/mailman/listinfo/python-list

We are using the LLNL module(s) on IRIX 6.5,
either with ABI mode -n32 or -64.
We met some problems with the Fortran calls we had in our
application, some of the Numeric object files were hidding
the native fortran library. We removed the f2c_lite.o object
file from the library we were using. This now works... perfectly.

Marcvs [alias Hop! Zizelp...]



From herzog at online.de  Tue Nov  9 11:01:30 1999
From: herzog at online.de (Bernhard Herzog)
Date: 09 Nov 1999 17:01:30 +0100
Subject: Graphics.
References: <3826E949.BBA1BAFD@imf.au.dk> <091119991052098844%kantel@mpiwg-berlin.mpg.de> <80973f$poq@newton.cc.rl.ac.uk>
Message-ID: <m3zownmz6t.fsf@greebo.nodomain.de>

"Richard Brodie" <R.Brodie at rl.ac.uk> writes:

> ...and as everyone knows, if you haven't done Python in German, you
> haven't done Python :)
> 
> Ich bin ein Holzf?ller und f?hl mich stark,
> Ich schlaf des Nachts und hack am Tag.

Shouldn't that have been

Ich bin ein Pyonier und f?hl mich stark,
Ich hack des Nachts und schlaf am Tag.

?


-- 
Bernhard Herzog	  | Sketch, a drawing program for Unix
herzog at online.de  | http://www.online.de/home/sketch/



From xfb52.no_spam at dial.pipex.com  Sat Nov  6 06:08:05 1999
From: xfb52.no_spam at dial.pipex.com (Alex Zbyslaw)
Date: Sat, 06 Nov 1999 11:08:05 +0000
Subject: Control-C on Unix misbehaviour?
References: <000101bf193a$1a2fc8c0$842d153f@tim> <m3emesdfjc.fsf@atrus.jesus.cam.ac.uk> <19991018154752.A16714@testserv.mathi.uni-heidelberg.de> <slrn80n2a1.5cr.quinn@seniti.ugcs.caltech.edu>
Message-ID: <38240C15.1D841E4@dial.pipex.com>

Quinn Dunkan wrote:
> 
> On Mon, 18 Oct 1999 15:47:52 +0200, flight at mathi.uni-heidelberg.de
> Here's a data point for Irix6:
> If you hit ^C at the >>> prompt with readline loaded (python 1.5), python
> stops responding.  Further ^Cs do nothing, in fact the only way I've found
> to get out is to kill from another shell or hit ^z kill %.

I had exactly the same problem on FreeBSD 3.2  but ONLY when I had
thread support enabled.

I have eventually tracked the problem down to a bad interaction (read
library/kernel bug) between threads/sigjmp and interrupted reads.

I have created two "fixes" which can be applied either to a) python or
b) python and readline which fix the problems differently.  The fixes
are pretty trivial and so far they work for me.  You can check them out
at http://www.xfb52.dial.pipex.com/patches/python.shtml

I don't know how they may affect people who have had ^C dumping them
into a shell.  No one else has mentioned whether they had thread support
enabled or not.

>From extensive poking around I am quite sure that this misbehaviour is
not a bug in either Python or readline.

Good luck,

--Alex



From mcfletch at vrtelecom.com  Tue Nov 16 15:26:04 1999
From: mcfletch at vrtelecom.com (Mike Fletcher)
Date: Tue, 16 Nov 1999 15:26:04 -0500
Subject: list.without()?
In-Reply-To: <38319104.F54AECEF@appliedbiometrics.com>
Message-ID: <NDBBIAPHEKGOBIHCBLIIEEOOCAAA.mcfletch@vrtelecom.com>

Hmm, actually, it's only cheap for certain subsets of the problem (which I
took to be the common case).  In particular, it shines for small numbers of
instances in lists of fairly arbitrary size.  It falls down completely when
you are removing large numbers of instances (such as all of them).  I just
tested five algos, testing 1 instance in middle of list, and a list composed
of only elements to be removed, findings, module, and raw test results
below...

Conclusions:
	First is the posted algo (while,try,except,break) <without> -- gives very
good performance for the few-in-many case but very poor performance for the
many in many cases for large values of many
	Second is a for loop with == test and append <without2> -- gives decent
(but not spectacular) results whenever memory effects are not a big factor,
not really a contender.
	Third is the lambda solution (which does surprisingly well) <without3> --
gives acceptable but not great performance all over, seems strange that
function-call-overhead is so minor a factor.
	Fourth is a backwards-index-counting scheme with del <without4> -- gives
marginally better performance than third on few-in-many case, marginally
worse than third in many-in-many case.
	Fifth is Preston's remove function <remove> -- gives best results for
few-in-many case, again, very poor results in many-in-many case

Summary:
	Looks like the fourth algo is likely the "winner" in this set (assuming
few-in-many is the most common case, but that many-in-many is a potential
case).  Third might be preferred due to the simplicity of the implementation
and similar speed.  For maximum few-in-many speed, the fifth is the most
appropriate.

Module is below, along with results run on a PIII 500 under NT.  Enjoy,
Mike

8<_____ without.py ____________________
def without( source, element):
	temp = source[:]
	while temp:
		try:
			temp.remove( element )
		except:
			break
	return temp

def without2( source, element ):
	temp = []
	for el in source:
		if el != element:
			temp.append( el )
	return temp

def without3( source, element ):
	return filter(lambda x, element=element: x!=element, source)
def without4( source, element ):
	temp = source[:]
	for index in xrange( len(temp)-1, -1, -1 ):
		if temp[index] == element:
			del temp[index]
	return temp
def remove(list, element):
    while element in list:
        list.remove(element)
    return list

def test( function):
	import time
	print "Testing function:", function
	print "  single element at middle:"
	for dl in (10, 100, 1000, 10000, 100000, 1000000):
		t = time.time()
		res = function( range( dl),  dl/2)
		delta = time.time()-t
		print "    %s: %s"%(dl,delta)
		if delta > 10:
			print "###Note: Tests skipped due to time overrun"
			break
	print
	print "  all elements being removed:"
	for dl in (10, 100, 1000, 10000, 100000, 1000000):
		t = time.time()
		res = function( [1]*dl,  1)
		delta = time.time()-t
		print "    %s: %s"%(dl,delta)
		if delta > 10:
			print "###Note: Tests skipped due to time overrun"
			break
	print


if __name__ == "__main__":
	#test( without )
	#test( without2 )
	test( without3 )
	test( without4 )
	test( remove )

8<_____ run results  ___________________
p:\>without
Testing function: <function without at 7d7f00>
  single element at middle:
    10: 0.0
    100: 0.0
    1000: 0.0
    10000: 0.00999999046326
    100000: 0.0699999332428
    1000000: 1.11199998856

  all elements being removed:
    10: 0.120000004768
    100: 0.0
    1000: 0.0100001096725
    10000: 0.770999908447
    100000: 78.7929999828
###Note: Tests skipped due to time overrun

Testing function: <function without2 at 7d7f80>
  single element at middle:
    10: 0.0
    100: 0.0
    1000: 0.0100001096725
    10000: 0.039999961853
    100000: 0.460999965668
    1000000: 19.6879999638
###Note: Tests skipped due to time overrun

  all elements being removed:
    10: 0.120000004768
    100: 0.0
    1000: 0.0
    10000: 0.0100001096725
    100000: 0.110999941826
    1000000: 1.16100001335

Testing function: <function without3 at 7d8f40>
  single element at middle:
    10: 0.0
    100: 0.0
    1000: 0.0
    10000: 0.0299999713898
    100000: 0.341000080109
    1000000: 3.39499998093

  all elements being removed:
    10: 0.129999995232
    100: 0.0
    1000: 0.0
    10000: 0.0299999713898
    100000: 0.320000052452
    1000000: 3.17499995232

Testing function: <function without4 at 7d8f60>
  single element at middle:
    10: 0.0
    100: 0.0
    1000: 0.0
    10000: 0.0199999809265
    100000: 0.19000005722
    1000000: 1.9129999876

  all elements being removed:
    10: 0.120000004768
    100: 0.0
    1000: 0.0
    10000: 0.039999961853
    100000: 0.401000022888
    1000000: 4.00499999523

Testing function: <function remove at 7daac0>
  single element at middle:
    10: 0.0
    100: 0.0
    1000: 0.0
    10000: 0.00999999046326
    100000: 0.0509999990463
    1000000: 0.580000042915

  all elements being removed:
    10: 0.130999922752
    100: 0.0
    1000: 0.00999999046326
    10000: 0.771000027657
    100000: 79.1130000353
###Note: Tests skipped due to time overrun




From ScherBi at BAM.com  Mon Nov 29 09:34:54 1999
From: ScherBi at BAM.com (ScherBi at BAM.com)
Date: Mon, 29 Nov 1999 09:34:54 -0500
Subject: Unbuffered stdout
Message-ID: <3F8BA119B48AD11197470008C7A493CB98B693@HQOBGEX1>

try using

#!/usr/bin/env python -u 

at the top of your scripts (assuming unix, or at least apache here)

see python --help for more info.

Bill



> -----Original Message-----
> From: leon [mailto:leonar_n at epita.fr]
> Sent: Monday, November 29, 1999 9:17 AM
> To: python-list at python.org
> Subject: Unbuffered stdout
> 
> 
> Hi there !
> 
> I'm developping an IRC like software which must work on the 
> Web. I can't use
> Java,
> and I'm using Python to program the CGI.
> But, I have a big problem : I don't know how to make the 
> stdout unbuffered.
> I tried to make some
>     sys.stdout.write("test")
>     sys.stdout.flush()
> but it don't work.
> 
> So any idea ??
> 
> 
> Leon.
> 
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 



From nemeth at iqsoft.hu  Wed Nov 17 05:46:08 1999
From: nemeth at iqsoft.hu (Nemeth Miklos)
Date: Wed, 17 Nov 1999 11:46:08 +0100
Subject: Beginning programming with Python [my last comment]
References: <sng0ychgwt.fsf@motorola.com> <382B3A53.5816B2@callware.com> <382B3D79.74CD3E0F@email.sps.mot.com> <80p9ds$ppo$1@newshost.accu.uu.nl> <asteroidal-80psfo/INN-2.2.1/bat@broadway.news.is-europe.net> <38312D55.A36EB56D@iqsoft.hu> <antigorite-80rol2/INN-2.2.1/attentive@broadway.news.is-europe.net> <383195C1.826B3342@iqsoft.hu> <allege-80s674/INN-2.2.1/apply@broadway.news.is-europe.net>
Message-ID: <38328770.EDFD9352@iqsoft.hu>

Thank you for all remarks, opinions about the topic. All replies helped me a
lot on my way to the Python World.
Yesterday night I had a determining experience. I was evaluating Fnorb (a
Python CORBA ORB implementation) and it worked perfectly, without a hitch
with my quite complex CORBA server written in C++ based on ORBacus. And even
more, writing CORBA applications in Python is extremelly simple and easy to
read, much easier than in Java or C++. This is a testimonial that CORBA
programming with Python is a joy. The CORBA support of Python is of a top
priority in my case, because the telecom corporation I work for decided CORBA
as the standard enterprise platform (aka bus).
I evaluated Perl, Tcl, and PHP but neither of them has acceptable CORBA
support (ie. extension modules).

Miklos Nemeth








From aahz at netcom.com  Mon Nov  1 08:33:30 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 1 Nov 1999 13:33:30 GMT
Subject: Newbie Question
References: <7vidsi$qm8$1@nnrp1.deja.com>
Message-ID: <7vk4ra$3hp$1@nntp6.atl.mindspring.net>

In article <7vidsi$qm8$1 at nnrp1.deja.com>,  <strangiato at my-deja.com> wrote:
>
>I am completely new to the language.

Congratulations!

>I heard that Yahoo uses it?  Is this true?  Is it a good language for
>developing Web Pages?  Is it a "best of breed"?

Yes.  Yes.  Yes.  Maybe.

Python is certainly my favorite language (currently), but I've used
enough languages over the years that I'm not going to make wildly
exaggerated claims for Python.  What I will claim is that all but
poorly-written Python code is much more readable than any other language
I've tried.
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From x at x.x  Wed Nov  3 19:33:48 1999
From: x at x.x (Eric Jacobs)
Date: Wed, 03 Nov 1999 19:33:48 -0500
Subject: Namespace question...
References: <7vqgib$l8v$1@nnrp1.deja.com>
Message-ID: <3820D46C.517EA278@x.x>

oofoe at my-deja.com wrote:
> 
> Hi!
> 
> How can you access the global namespace in python 1.5.1? I'd like to get
> a version string which is declared in the main script file. However,
> each module also has a version string which overrides the original one.
> I tried to access it as __main__.Version, similarly to $main::Version in
> Perl, which is what I'm more familiar with. Of course, it didn't work.
> 
> I have something like this:
> 
> module file foo.py:
> ------------------------------
> Version = '$Id$'
> 
> def bar():
>    print "Version is: ", __main__.Version
> 
> main program:
> ------------------------------
> Version = '$Id$'
> 
> import foo
> 
> foo.bar()
> 
> -------------------------------
> 
> I'd certainly appreciate any pointers. The namespace stuff seems to get
> a little hazy around issues like this.

Import it first! Use import __main__.

I think some of this confusion arises because there is no global
namespace in the sense of a area that exists above all modules
(a good design design, IMO). The word global simply refers to 
variables at the level of the module, as opposed to within a
function.



From gmcm at hypernet.com  Mon Nov  1 07:32:58 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Mon, 1 Nov 1999 07:32:58 -0500
Subject: Comparison between Python and "Ruby"
In-Reply-To: <7vjtrc$hmm$1@newshost.accu.uu.nl>
Message-ID: <1270663667-24391282@hypernet.com>

Martijn Faassen asks:

> What-does-TMTOWTDI-stand-for?-ly yours,

That's the Perler's war cry. Snort. As if there was only one 
way to have more than one way to do it.

(As used by Matz, of course, it does not qualify as an 
oxymoron.)

- Gordon



From schliep at Octopussy.MI.Uni-Koeln.DE  Fri Nov  5 23:06:22 1999
From: schliep at Octopussy.MI.Uni-Koeln.DE (Alexander Schliep)
Date: 06 Nov 1999 05:06:22 +0100
Subject: Drawing (directed) graphs
References: <ye1bt9das2p.fsf@ex.ac.uk>
Message-ID: <tq9d7top8ld.fsf@Octopussy.MI.Uni-Koeln.DE>

R.M.Everson at ex.ac.uk (R.M.Everson) writes:

> Hello,
> 
> Does anyone know of Python code for drawing graphs of the
> graph-theoretic type? Ideally, I'd like to be able to interactively
> add nodes and edges etc and add data to structures associated with the
> nodes and edges.

Checkout http://www.zpr.uni-koeln.de/~gato. Gato comes with gred the graph
editor. It does not do automatic layout though.

Yours,
Alexander


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



From michelorengo at netscape.com  Tue Nov  2 10:49:33 1999
From: michelorengo at netscape.com (Michel Orengo)
Date: Tue, 02 Nov 1999 15:49:33 GMT
Subject: Installer v0.3d - problem with Builder
Message-ID: <381F0803.6D76E236@netscape.com>

I've tried to use the new beta of installer (simple) but I've got an
Attribute Error at line 109 in 'archmodule.modules'
With installer v0.2 (simple), it works fine.
Any idea of what went wrong?
Thanks for your help

Michel Orengo




From tim_one at email.msn.com  Mon Nov  8 12:03:31 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Mon, 8 Nov 1999 12:03:31 -0500
Subject: __str__ vs. __repr__
In-Reply-To: <u4glOIWXzGrqvJnI+IlPAQeurz1J@4ax.com>
Message-ID: <000101bf2a0b$2f87d0a0$e1a2143f@tim>

>> repr(obj) should return a string such that
>>    eval(repr(obj)) == obj

[Toby Dickenson]
> I don't understand the motivation for this requirement. Why would
> anyone want to pass such a string to eval? If you anticipate the need
> for reconstructing the object from a textual representation, then
> surely pickle is a better option?

This was well covered by others (not everything of common interest can be
pickled; and pickle strings, while portable, are wholly unreadable by
people).

> ...
> OK, so repr as it stands today is clearly not suitable for formatting
> for people,

It's much better suited for that than is pickle.  That is, people can and do
cut & paste reprs with confidence, and even edit them by hand.  So, indeed,
the requirement at the top of the msg doesn't capture *everything* a good
repr should do.

> and is less than ideal for formatting for the machine.

Based on the repr(float) example?  That one is easy to repair.  So long as
interactive mode uses repr() for raw expressions, though, and containers
don't "pass str() down", most people won't want to see the "extra" digits
most of the time.

> However there is a third audience for who it is ideal, the programmer,
> who need a representation for use in development tools such as
> debuggers, or the interactive mode.

People (no, not even programmers <wink>) aren't that monolithic; e.g., in
development tools it's likely you're going to want to see the repr() of your
own objects and the str() of mine.  I'm not trying to address that --
there's only so much we can squeeze out of two functions.  It would be a
real help if those two (repr & str) had clearly understood purposes,
though -- they would cover more of the space if they weren't treated like
fuzzy synonyms in practice.

> Tim's Rat class is a good example of how __repr__ should not be
> implemented, because it exposes the internal representation of the
> object, not the external value.

Beg to differ:  it's an excellent example of how __repr__ should be
implemented.  The information in a *typical* Rat can require terabytes of
string space to represent if expressed as, e.g., a decimal expansion.  It's
a case where "the internal representation" is no accident:  a pair of longs
is one of a very few tractable ways to capture the object's true value.
Approximations are fine for __str__, though (and, although I didn't show it,
Rat.__str__ supports a dozen options for making nice-looking and compact
approximation strings -- alas, it's hard to talk Python into *invoking*
Rat.__str__!).

> However, using a small number of digits for repr(float) does make
> sense because it still provides sufficient precision for most
> programmers.

See above:  I told you you'd want to see the str() of everyone's objects
except your own <0.5 wink>.  This isn't a joke to people who use floats
seriously -- just as you take your objects seriously.

most-americans-don't-give-a-rip-about-o-umlaut-either-ly y'rs  - tim





From amitp at Xenon.Stanford.EDU  Wed Nov 24 16:12:13 1999
From: amitp at Xenon.Stanford.EDU (Amit Patel)
Date: 24 Nov 1999 21:12:13 GMT
Subject: Help with SMTP
References: <81hase$bdj$1@news.fsu.edu>
Message-ID: <81hkbd$1m8$1@nntp.Stanford.EDU>

 Glenn Kidd <gek8365 at garnet.acns.fsu.edu> wrote:
| Hello,
| 
| I wrote a few weeks ago and I received a lot of responses.  I have a new
| problem.  I am trying to write a form mailer and I have gotten the script to
| accept the form data and mail the info.  The message is received and
| everyhting looks in order but the website reports a "500 Internal Server
| Error".  When I look at the server log I get the following description of
| the error:

Is your CGI script printing anything out?  It needs to print out some
headers and a web page body (in HTML format, most likely).

	- Amit




From timmuddletin at news.vex.net  Fri Nov 26 22:55:48 1999
From: timmuddletin at news.vex.net (tim muddletin)
Date: 27 Nov 1999 03:55:48 GMT
Subject: Sending files to a WWW-Server
References: <81nje2$su9$1@rznews.rrze.uni-erlangen.de>
Message-ID: <81nko4$18hg$1@hub.org>

On Sat, 27 Nov 1999 04:32:33 +0100, Jan Stiller wrote:
>I'm searching for examples how to send files to a web server. I've found

Two ever popular examples listed here:
http://www.vex.net/~x/parnassus/apyllo.py?so=d&find=cgi+upload





From hniksic at iskon.hr  Tue Nov  2 10:41:03 1999
From: hniksic at iskon.hr (Hrvoje Niksic)
Date: 02 Nov 1999 16:41:03 +0100
Subject: Printing of individually returned values when running interactively
Message-ID: <9t966zk6gvk.fsf@mraz.iskon.hr>

I've just noticed (the hard way) a very unusual feature of Python's
interactive interpreter mode.  If you evaluate several values on
several lines, all the values are printed!  For instance:

>>> if 1:
...  1
...  2
...  3
... 
1
2
3

My remark will probably look strange to seasoned Pythoneers, but I'm
really surprised by discovering this.  Why would anyone want a thing
like that?  Doesn't it produce loads of debug-like output when pasting
more complex things?

I guess I'm more used to Lisp interactive interpreters whose REPL is
sexp-oriented, and prints the value of the last read sexp.


(If you're curious, the case that hit me hard was when I tested
mxODBC's connection with a remote database at work.  I wrote something
like:

for i in range(10000):
  cur.execute('insert into test ...', i)

Needless to mention, the execute() method returns a number!)



From muniff at yahoo.com  Thu Nov 25 03:15:23 1999
From: muniff at yahoo.com (Ahmad Munif)
Date: Thu, 25 Nov 1999 08:15:23 GMT
Subject: how to split the contain of retr (using poplib module)
Message-ID: <383ceefd.3147421@news.dnet.net.id>

i'm using poplib module to connect to pop3 server.
i can retrieve mail with retr mothode.
but i don't know, how to split the contain of mail,
so, i can get the value of mail, such as: To,From,body message.
how to split it ?




From tim_one at email.msn.com  Sat Nov  6 04:02:47 1999
From: tim_one at email.msn.com (Tim Peters)
Date: Sat, 6 Nov 1999 04:02:47 -0500
Subject: __str__ vs. __repr__
In-Reply-To: <7vvni7$19q0$1@nntp6.u.washington.edu>
Message-ID: <000201bf2835$b2c9a380$9c2d153f@tim>

[Tim]
>> repr(obj) should return a string such that
>>     eval(repr(obj)) == obj

[Donn Cave]
> For real?

This, along with the later claim about what str() "should do", are
paraphrased from the Language Reference Manual (3.3.1 Basic customization).
It's "why" e.g. the equation above holds for repr(string) but not for
str(string).  The builtin types took this stuff as seriously as they could,
modulo some unfortunate lapses (like repr(float)) covered elsewhere in this
thread.

>  I'm very used to seeing repr's that don't eval at all -
>
> '<socket object, fd=3, family=2, type=1, protocol=0>'
> "<open file '<stdout>', mode 'w' at 80005660>"
> '<exceptions.NameError instance at 80048ae0>'

Are you for real <0.9 wink>?  Be reasonable.

> ...
> Well, really I guess I would prefer that authors _not_ try to capture
> *everything* important, if that means a large data dump.  repr is used
> in lots of places, like the interactive interpreter, where the output
> only has to be roughly indicative,

The latter is str()'s purpose (see 3.3.1 again), and the thread was arguing
(among other things) that the interpreter use it (instead of repr())
interactively.

> and I have never seen and can't find any example of its use with
> eval or in any context where it has to be definitive.

I've seen it used often as a poor-man's pickle, and as a way to email Python
objects as plain text.  repr() is very useful for that.  Unfortunately, most
class authors seem to ignore the distinction (my classes don't!), and that
limits its usefulness.  The lapses in the core also hurt.  One email
response I didn't see posted here describes a typical example of the latter:

<ANONYMOUS>

I've only been writing Python for a year or so, but bumped into this 'cheat'
and had to write floats out to a file using binary format, a portability
nightmare.  The floats were in an array, and you can't pickle arrays.

>>> repr(12345678901234567.0)
'1.23456789012e+016'  # Just 12 digits!

is currently what you get.

>>> x = 12345678901234567.0
>>> x == eval(repr(x))  # Nope.
0

It sure would be nice to be able to count on
    eval(repr(obj)) == obj.

On the other hand, while repr(1L) has to return '1L',  there's no reason
str(1L) shouldn't be '1'.

</ANONYMOUS>

OTOH,

>>> x == eval("%24.16e" % x)
1
>>>

on any IEEE-754 conforming platform, as was covered in detail earlier.
That's what repr(float) should do.

> If there are folks who want to provide this functionality at the expense
> of making a string that's unusable for repr's normal applications,
> they should put it in a different function.

The language took its stand on this issue in 1991 <wink>, and you're arguing
against *that*.  I'm arguing for the language to finish what it started.

> ...
> Bah.  If we get str() because it's pleasant to read, aren't we worried
> that repr() will be harsh and stressful?  Will one repr() too many send
> one of us into a homicidal rage one of these days?

I said before that repr() is rarely appropriate -- but essential when it is.

> Let us think about taming the cruel repr()!  If I may speculate on what
> you're trying to say, str() is pleasant because it shows the data in
> form that has been digested for human consumption.  In contrast to
> repr(), which is intended for the machine.

Ah -- you *have* read section 3.3.1 <wink>.

> ...
> So look at the way it's really used.

Inconsistently, confusingly, and inappropriately, in both classes and (but
substantially less so) in the core types.  We can't deduce anything useful
about what should happen now based on what is a mess in practice.  My aim is
to get the mess cleaned up.

and-thanks-for-your-support<wink>-ly y'rs  - tim





From fredrik at pythonware.com  Mon Nov 29 06:01:05 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 29 Nov 1999 12:01:05 +0100
Subject: Problems Finding and installing Tkinter
References: <383ED44D.3E31C959@dimensional.com>
Message-ID: <01bd01bf3a59$093a9dc0$f29b12c2@secret.pythonware.com>

Jim Cochrane <jtc at dimensional.com> wrote:
>     import _tkinter # If this fails your Python is not configured for Tk
> 
> ImportError: /usr/X11R6/lib/libtk8.0.so: undefined symbol: panic
> 
> So it's not finding a symbol named "panic".  I have no idea what library
> this should be in.

libtcl8.0.so, most likely:

    $ nm libtk8.0.so | grep panic
             U panic
    $ nm libtcl8.0.so | grep panic
    0000afac T panic
    0005af58 D panicProc

</F>

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




From spamfranke at bigfoot.de  Tue Nov  2 13:19:14 1999
From: spamfranke at bigfoot.de (Stefan Franke)
Date: Tue, 02 Nov 1999 18:19:14 GMT
Subject: Drawing (directed) graphs
References: <ye1bt9das2p.fsf@ex.ac.uk>
Message-ID: <381f2a3b.27625609@news.btx.dtag.de>

There's a package for the wxWindows GUI toolkit called OGL (object
graphics library). A demo is included in wxPython.

Very simple - you can only place & resize vertices with the mouse and
the edges follow -, but could be a start.

Stefan


On Tue, 2 Nov 1999 14:25:34 GMT, R.M.Everson at ex.ac.uk (R.M.Everson)
wrote:

>
>Hello,
>
>Does anyone know of Python code for drawing graphs of the
>graph-theoretic type? Ideally, I'd like to be able to interactively
>add nodes and edges etc and add data to structures associated with the
>nodes and edges.
>
>Many thanks
>Richard Everson.




From andres at corrada.com  Sun Nov 21 11:50:11 1999
From: andres at corrada.com (Andres Corrada)
Date: Sun, 21 Nov 1999 11:50:11 -0500
Subject: Order of keyword arguments
References: <38379233.ADE65CDB@corrada.com> <19991121034245.A21255@quark.emich.edu> <38380F92.D9817451@corrada.com> <19991121114827.A21697@quark.emich.edu>
Message-ID: <383822C3.DD0CE177@corrada.com>

Jeff wrote:

> ....................................................... but I don't see the
> advantage to 'knowing' the order that the keywords were used by the calling
> function.
> 

I think there would be an advantage to having such a data structure as
an ordered dictionary. My "American Heritage Dictionary of the English
Language" is such a data structure.

------------------------------------------------------
Andres Corrada-Emmanuel   Email: andres at corrada.com
Owner                     http://www.corrada.com/mamey
Mamey                     Phone: (413) 587-9595
------------------------------------------------------



From arcege at shore.net  Thu Nov  4 16:47:53 1999
From: arcege at shore.net (Michael P. Reilly)
Date: Thu, 04 Nov 1999 21:47:53 GMT
Subject: C-Python tutorial?
References: <19991101203120.A29797@optiplex.palga.uucp>
Message-ID: <danU3.90$AT.20506@news.shore.net>

Gerrit Holl <gerrit.holl at pobox.com> wrote:
: Hello,

: I have learned python as my first langauge some time ago. Is there a C
: tutorial for python programmers, that covers C programming, espaccialy for
: extending it with python?

: I think not. Can anyone write one? I would be very grateful!

There is not C _tutorial_, but a documents exist on how to write C
extensions for Python, embed Python in C and to write your own C types
for Python.

The first two are on the Python website and the last is at:
  http://starship.python.net/~arcege/extwriting/pyext.html

When I get some more time here at my new job, I'll flesh this doc out,
get it formatted and sent off to Fred.

  -Arcege




From jon_mclin at attglobal.net  Tue Nov 30 06:42:14 1999
From: jon_mclin at attglobal.net (Jon McLin)
Date: Tue, 30 Nov 1999 11:42:14 +0000
Subject: Installation Problem (1.5.2)
Message-ID: <3843B816.F4A31F07@attglobal.net>

On my Mandrake 6.0 box, I just upgraded from 1.5.1 to 1.5.2 with a fresh
compile.  Since the 1.5.1 installation placed the binaries in /usr/bin, I used
the configure option "--exec_prefix=/usr"

My procedure:
rm configure.cache
make clean
./configure --exec_prefix=/usr
make OPT=-O0        # optimization off due to compiler problem (else fails
test_long, test_types)
make test
make install

Unfortunately, when I subsequently start the interpreter in /usr/bin, I get:

Built-in exception class not found: EnvironmentError.  Library mismatch?
Warning!  Falling back to string-based exceptions
Python 1.5.2 (#7, Nov 30 1999, 04:22:04)  [GCC pgcc-2.91.66 19990314
(egcs-1.1.2  on linux2

The interpreter in the build directory starts successfully (with ./python):

Python 1.5.2 (#7, Nov 30 1999, 04:22:04)  [GCC pgcc-2.91.66 19990314
(egcs-1.1.2  on linux2
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam

If I build and install with the default (to /usr/local/), /usr/local/bin/python
starts successfully.

No python variables are present in my environment.

What am I missing?

Thanks,
Jon





From boud at rempt.xs4all.nl  Fri Nov 19 05:53:45 1999
From: boud at rempt.xs4all.nl (Boudewijn Rempt)
Date: Fri, 19 Nov 1999 11:53:45 +0100 (CET)
Subject: Beginning programming with Python
In-Reply-To: <012401bf3277$72897010$f29b12c2@secret.pythonware.com>
Message-ID: <Pine.LNX.4.10.9911191150240.2571-100000@essle.valdyas.org>

On Fri, 19 Nov 1999, Fredrik Lundh wrote:

> Boudewijn Rempt <boud at rempt.xs4all.nl> wrote:
> > You needn't guess about that, I guess ;-). I didn't know there
> > was a design team for Python, I though it was Guido's invention.
> 
> it is -- the CWI researchers (including GvR) worked
> on Python's predecessor ABC.
> 
> > But I still wonder about the double == - if I can't assign in
> > an if, why insist on ==?
> 
> probably because binding and comparision
> are two different things; consider:
> 
>     a = b = c
> 
> vs.
> 
>     a = b == c
> 

Ah, that makes it a lot clearer - I just didn't know about a = b ==
c. After having earned the monthly envelope with software development
for a few years, I'm only know realizing how much I don't know, 
never having been educated for it. But I've started to read up on it!

Boudewijn Rempt  | http://denden.conlang.org




From mwh21 at cam.ac.uk  Wed Nov 24 12:23:38 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 24 Nov 1999 17:23:38 +0000
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <8666ysfta9.fsf@ix.netcom.com> <012e01bf365d$912fa360$f29b12c2@secret.pythonware.com> <5g08OLe6ncyc5Uu5627KDJVZ2N0R@4ax.com>
Message-ID: <m3wvr7st0l.fsf@atrus.jesus.cam.ac.uk>

Dennis Lee Bieber <wlfraed at ix.netcom.com> writes:

> On Wed, 24 Nov 1999 10:23:27 +0100, "Fredrik Lundh"
> <fredrik at pythonware.com> declaimed the following in comp.lang.python:
> 
> > Brad Knotwell <knotwell at ix.netcom.com> wrote:
> > > Similarly, the only thing I've found missing from Python is C's
> > > ternary operator, it would be nice to be able to do something like 
> > > the following:
> > > 
> > > def max(x,y): return ((x>y) ? x : y)
> > 
> > now that was one lousy example ;-)
> >
> 	Also potentially unsafe if either x or y is some expression when
> invoked. At least, I seem to recall some suggestion to always put macro
> arguments in parens...
> 
> 	def max(x,y): return(((x)>(y)) ? (x) : (y))

Eh? This is Python, not some yucky character-based macro equipped UNIX
implementation language <wink>.

One thing that 

x = (var and (ifexp,) or (elseexp,))[0]

has over

def choose(var,ifvar,elsevar):
    if var:
        return ifvar
    else:
        return elsevar

x = choose(var,ifexp,elseexp)

besides brevity is that the former will only evaluate one of ifexp or
elseexp, but that isn't going to matter for max (obviously...).

It can be handy in things like:

x = (hasattr(x,"foo") and (x.foo,) or (None,))[0]

though in this case 

x = getattr(x,"foo",None)

is clearly better...

Regards,
Michael



From foxes at theriver.com  Tue Nov  2 14:10:47 1999
From: foxes at theriver.com (Andy Fox)
Date: Tue, 02 Nov 1999 12:10:47 -0700
Subject: How to control app from Python?
References: <38160DD8.5C695E09@theriver.com> <3817BD5F.7E05BDC9@home.com> <38185A03.9AB0FAD6@theriver.com> <7vke68$6r6$1@nnrp1.deja.com> <7vkvia$st7@mail.psy.uva.nl>
Message-ID: <381F3737.3D20B5E9@theriver.com>

ionel,

Thanks for the code example.  This looks like _exactly_ what I'm trying
to do with Python.  The app I want to control is neither a DDE nor COM
server.

Thanks again,
-- 
Andy Fox



From tottinge at concentric.net  Tue Nov 30 13:59:41 1999
From: tottinge at concentric.net (Tim Ottinger)
Date: 30 Nov 1999 10:59:41 PST
Subject: WinNT and python as text filter...
References: <3842f32d.30322451@news.concentric.net> <blQ04.11116$VA6.52672@news-server.bigpond.net.au>
Message-ID: <38442033.107384791@news.concentric.net>

On Tue, 30 Nov 1999 13:31:51 GMT, "Mark Hammond"
<mhammond at skippinet.com.au> wrote:

>How are you executing the code?  Are you just typing:
>
>c:\>myscript.py > out
>
>If so, there is a bug in the NT command prompt - you need to specify the
>full path to Python.exe on the command line, and it will work...
>

This works:
	c:\> myscript.py
and produces the output I want. This, however:
        c:\> myscript.py > somefile.txt
produces an empty text file.

I have somewhat better luck with:
        c:\> python myscript.py > somefile.txt

I want this to work from within vim, so it's pretty important that I
have pipeable output. Of course, better yet would be getting the
python stuff compiled into my vim....

tim




From preston at jump.net  Wed Nov 17 22:22:08 1999
From: preston at jump.net (Preston Landers)
Date: 17 Nov 1999 21:22:08 -0600
Subject: How to capture stdout in an exec'ed string?
References: <80v7j4$g0n$1@nnrp1.deja.com> <pbGY3.21$VA6.239@news-server.bigpond.net.au>
Message-ID: <87bt8sqyan.fsf@gandalf.localnet>

Thanks a lot Mark!  It only took a few microseconds after I posted to
realize what I needed to do.  For some reason the concept of saving
and restoring the stdout handler didn't occur to me.  I was thinking
about namespace issues which are not really related.

---Preston

"Mark Hammond" <mhammond at skippinet.com.au> writes:

> Something like this should work:
> 
> try:
>    oldout = sys.stdout
>    sys.stdout = My_STDOUT()
>    try:
>      cruft = my_ancient_class()
>      cruft.do_your_thang()
>    finally:
>      sys.stdout = oldout
> except SystemExit, msg:
>    process(msg)
> 
> 
> Mark.



From nhv at cape.com  Thu Nov 25 15:36:45 1999
From: nhv at cape.com (Norman Vine)
Date: Thu, 25 Nov 1999 15:36:45 -0500
Subject: Journaling Code
Message-ID: <81k6m9$gkm$1@newsie2.cent.net>

I am involved with a new Open Source 3D Modeller project
and we are investigating various undo-redo mechanism's.

A member of the group suggested implementing a scripting
layer between the modeller and the GUI, and journalling all
user invoked scripts in order to be able to rebuild the model
from a saved state by simply replaying the journaled code.

I am wondering if anyone has used Python for anything similar,
and if so if they would be willing to share their knowledge.

Cheers

Norman Vine

archives  http://www.topica.com/lists/3dmodeller
URL ; http://www.woodsoup.org/~sbaker/modellers







From mhammond at skippinet.com.au  Sat Nov 27 22:03:22 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sun, 28 Nov 1999 03:03:22 GMT
Subject: Pythonwin Reset
References: <01bf38df$8127e0c0$3a42c3d1@Cyberus.cyberus.ca>
Message-ID: <_X004.7432$VA6.38559@news-server.bigpond.net.au>

Nope.  The closes Pythonwin could really do is to perform a
__main__.__dict__.clear(), and that is pretty easy to type.  There is no way
Pythonwin could perform a more intrusive reset, as Pythonwin is written in
Python!

Mark.

Gordon Williams wrote in message
<01bf38df$8127e0c0$3a42c3d1 at Cyberus.cyberus.ca>...
>Hello All,
>
>Is there a way to reset Pythonwin IDE between runs.  I want to eliminate
>any rogue objects or modules that are still in memory from the previous run
>of the program.
>
>Thanks
>
>Gordon Williams





From greg.ewing at compaq.com  Thu Nov 25 05:57:13 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Thu, 25 Nov 1999 23:57:13 +1300
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <867lj85g4c.fsf@g.local>
Message-ID: <383D1609.322062A5@compaq.com>

Gareth McCaughan wrote:
> 
>     if you adopt an OO approach to the
>     parse tree (with things like "class IfStatement(ParseNode):")
>     then you can't separate out very different operations like
>     foo.print(), foo.compile() and foo.evaluate(). (That is,
>     you can't group all the print methods together, and group
>     all the compile methods together somewhere else, etc.)

There is a way to do this. Instead of defining just one class
for each parse node, you define classes PrintIfStatement,
CompileIfStatement, EvaluateIfStatement, etc. which contain
the appropriate methods. Then you define the main IfStatement
class so that it inherits all of these.

The individual classes can then be grouped however you want,
e.g. you could put all the Compile* classes in a Compile
module, etc.

Greg



From thantos at chancel.org  Wed Nov 24 08:58:26 1999
From: thantos at chancel.org (Alexander Williams)
Date: Wed, 24 Nov 1999 13:58:26 GMT
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <383979F7.25D6@cs.bham.ac.uk>  <81fsab$2b$1@news.udel.edu>
Message-ID: <slrn83nrva.7gk.thantos@chancel.org>

On 24 Nov 1999 05:15:55 GMT, Terry Reedy <tjreedy at udel.edu> wrote:
>You are currently welcome to save 3 keystrokes (rather than 4)
>by naming the method parameter 's' instead of 'self'.  The latter

In a fit of lazy genius, I struck upon using 'me' instead; its a more
active referent /and/ faster to type than 'self.'  :)

-- 
Alexander Williams (thantos at gw.total-web.net)
"In the end ... Oblivion Always Wins."



From cfelling at iae.nl  Sat Nov  6 12:37:42 1999
From: cfelling at iae.nl (Carel Fellinger)
Date: 6 Nov 1999 18:37:42 +0100
Subject: making the Python case to my dept.
References: <9911051740130D.02530@william.jinsa.org> <7GOU3.272$Ae3.194028@news-dal.corridex.com>
Message-ID: <801p16$10s$1@vvs.superst.iae.nl>

Evan Simpson <evan at tokenexchange.com> wrote:

> Now I'm preparing to introduce my children to programming through Python.
> I'm starting with their fascination with Alice, the 3-D animation workshop
> which even my 6 year old enjoys playing with (though he doesn't script it...
> yet).

where can I find more on this Alice thing? my kids would love to play with
3-D animations! (I take it that scripting it is with Python)

thanks for the pointer

-- 
groetjes, carel



From bernhard at alpha1.csd.uwm.edu  Fri Nov 26 13:01:23 1999
From: bernhard at alpha1.csd.uwm.edu (Bernhard Reiter)
Date: 26 Nov 1999 18:01:23 GMT
Subject: NEWBIE
References: <81m7m3$82$1@news.qub.ac.uk>
Message-ID: <slrn83tinj.k7h.bernhard@alpha1.csd.uwm.edu>

On 26 Nov 1999 15:06:43 GMT, Stuarty at excite.co.uk <Stuarty at excite.co.uk> wrote:
>I have just recently started using Python, after the first day I was further 
>on than I am after a semester of C programming(were at arrays...).

Welcome to the pythondome. :)

>tutorial by Giudo van Rossum was very useful, but I attempted printing out the 
>reference library..there are now less considerably less trees in the world.

There are trees wasted and trees put to work.
	Bernhard
-- 
Research Assistant, Geog Dept UM-Milwaukee, USA.  (www.uwm.edu/~bernhard)
Free Software Projects and Consulting 		         (intevation.net)  
Association for a Free Informational Infrastructure            (ffii.org)



From philh at vision25.demon.co.uk  Mon Nov 29 18:25:54 1999
From: philh at vision25.demon.co.uk (Phil Hunt)
Date: Mon, 29 Nov 99 23:25:54 GMT
Subject: performance of tuple-less Python?
References: <1268226961-30285235@hypernet.com>
Message-ID: <943917954snz@vision25.demon.co.uk>

In article <1268226961-30285235 at hypernet.com>
           gmcm at hypernet.com "Gordon McMillan" writes:
> They're only redundant if the concept of mutability is 
> meaningless. A reference-based language without a concept 
> of mutability would be, um, an adventure.

Why?

-- 
Phil Hunt - - - phil at comuno.com
"Software is like sex, it's better when it's free" -- Linus Torvalds




From alex at somewhere.round.here  Tue Nov  2 13:41:15 1999
From: alex at somewhere.round.here (Alex)
Date: 02 Nov 1999 13:41:15 -0500
Subject: Printing strings
References: <381F00E1.721BB47F@cornell.edu>
Message-ID: <etdk8o0rb1w.fsf@w20-575-100.mit.edu>

> I'm trying to do something like the following:
> 
> a=4.0
> b=2.0
> print '2(%3.1f)' % a,b
> 
> where I really mean
> 
> print '%3.1f %3.1f' % a,b

You could use

print string.join(2*['%3.1f'])%(a,b)

I suppsoe.

Alex.



From bbosware at vic.bigpond.net.au  Mon Nov 29 23:37:34 1999
From: bbosware at vic.bigpond.net.au (John Leach)
Date: Tue, 30 Nov 1999 04:37:34 GMT
Subject: Looking for sshlib.py
Message-ID: <iwI04.9621$VA6.47863@news-server.bigpond.net.au>

Is there an ssh lib for Python. Or anything similar. I'm looking for
something more secure than telnet.
Any ideas? Write my own?
John Leach
http://www.bosware.com.au







From gerrit.holl at pobox.com  Fri Nov 19 07:34:04 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Fri, 19 Nov 1999 13:34:04 +0100
Subject: urllib.getsize()
In-Reply-To: <14388.29839.294164.707388@dolphin.mojam.com>
References: <19991118213925.A10221@optiplex.palga.uucp> <14388.29839.294164.707388@dolphin.mojam.com>
Message-ID: <19991119133404.A3308@optiplex.palga.uucp>

Skip Montanaro wrote:
> 
>     Gerrit> I've written a function to get the size of a HTTP/FTP url, is it
>     Gerrit> good enough to be put in urllib.py?
> 
> My only worry is that HEAD doesn't always work (for whatever reason having
> to do more with servers than clients).  Would it be better to wrap things in
> a try/except block?

Oh. Yes, I think getsize() should return -1 or None or an error if HEAD
doesn't work. Just doing a GET then isn't good, of course ;-)
Another problem: does every HTTP server give the Content-Length header?

> Anyone have a URL that fails with HEAD?

I don't have one.

regards,
Gerrit.

-- 
"We all know Linux is great...it does infinite loops in 5 seconds."
(Linus Torvalds about the superiority of Linux on the Amterdam
Linux Symposium)



From bjorn at roguewave.com  Fri Nov  5 12:57:14 1999
From: bjorn at roguewave.com (Bjorn Pettersen)
Date: Fri, 5 Nov 1999 10:57:14 -0700 
Subject: Comparison between Python and "Ruby"
Message-ID: <51C9F3C392D0D111A15600A0C99E64A9AF5629@localhost>

> Neel Krishnaswami wrote:
> > 
> > Python (and Ruby, for that matter) remind me more of Smalltalk than
> > anything else -- ie, they have extensive libraries, a 
> community with a
> > love of elegance but not of purity[*], and a greater focus 
> on practical
> > applications than on academic research.
> 
> Isn't the Smalltalk community pretty enamored of purity? I 
> mean "if" is
> an object that takes messages, right?
> 
>  Paul Prescod

Almost. A comparison creates a boolean object that takes ifTrue: and
ifFalse: (and whileTrue: etc.) messages.  Quite elegant...

-bjorn



From mhammond at skippinet.com.au  Mon Nov  1 05:57:58 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Mon, 1 Nov 1999 21:57:58 +1100
Subject: PythonWin Debugger
In-Reply-To: <001f01bf244c$43f705c0$3acbd9c2@peridot.optichrome.com>
Message-ID: <002401bf2457$f6500fe0$0501a8c0@bobcat>

> > Hmmm.  The debugger should run from all embedded apps,
> except Tkinter
> > ones (and therefore except IDLE).
>
> Does that include straight win32 (non-MFC) apps?

Yes - it should work there.  Eg, if you run
"pythonwin\pywin\debugger\fail.py" as a normal script from Python.exe,
the debugger should fire up correctly.

> > This particular problem would appear to be a bug.  I have another
> > debugger bug I need to track down, and hopefully I can also solve
> > this.
>
> Is it connected with this one, or is it completely different?

Different - occasionally the debugger seems to "hang" (although it
hasnt really - it just doesnt respond - as if a modal dialog is up,
although the modal dialog is hidden).  I can't repro this at the
moment, even tho I have seen it before :-(

> > OTOH, the best stragegy for this is probably remote
> debugging - use a
> > seperate process to perform the debugging.  This would also solve
a
> > few of the hairy "blocking" problems I need to deal with.
However,
> > this doesnt appear trivial either :-(
>
> I should imagine not. I suppose there is some sort of API
> under Windows
> to allow this sort of thing (like the Visual C++ debugger does) but
I
> wouldn't have any idea where to start.

That API is too low-level for Python to use effectively.  We really
dont need the ability to replace an instruction with a break-point.

Mark.




From nick at src.uchicago.edu  Tue Nov  2 15:50:57 1999
From: nick at src.uchicago.edu (Nick Collier)
Date: Tue, 02 Nov 1999 14:50:57 -0600
Subject: Can't start IDLE
Message-ID: <381F4EB1.AFD9BA32@src.uchicago.edu>

On winNT sp3 I get the following message when I do a "python idle.py"
from the command line:

C:\Program Files\Python\Tools\idle>python idle.py
Traceback (innermost last):
  File "idle.py", line 3, in ?
    PyShell.main()
  File "PyShell.py", line 612, in main
    fixwordbreaks(root)
  File "EditorWindow.py", line 606, in fixwordbreaks
    tk.call('tcl_wordBreakAfter', 'a b', 0) # make sure word.tcl is
loaded
TclError: invalid command name "tcl_wordBreakAfter"

I looked around deja.com and found one other person with the same
problem, but no response so I'm asking here.

thanks,

Nick

-- 

Nick Collier
nick at src.uchicago.edu
Social Science Research Computing
University of Chicago
Chicago, IL 60637



From jeremy at cnri.reston.va.us  Tue Nov 16 15:37:09 1999
From: jeremy at cnri.reston.va.us (Jeremy Hylton)
Date: Tue, 16 Nov 1999 15:37:09 -0500 (EST)
Subject: Final classes like in java???
In-Reply-To: <14385.48747.331513.47679@weyr.cnri.reston.va.us>
References: <80sao7$1848$1@rtpnews.raleigh.ibm.com>
	<14385.45862.55010.146256@weyr.cnri.reston.va.us>
	<14385.48236.202756.297241@goon.cnri.reston.va.us>
	<14385.48747.331513.47679@weyr.cnri.reston.va.us>
Message-ID: <14385.49269.629053.250222@goon.cnri.reston.va.us>

>>>>> "FLD" == Fred L Drake, <fdrake at acm.org> writes:

  FLD> I wrote:
  >> Not guaranteed, but this should be good enough:

  FLD> Jeremy Hylton writes:
  >> This doesn't really work.

  FLD>   That's exactly why I said it wasn't guaranteed.  ;-)

Would you be interested in buying my general purposes NP algorithm
solver?  It provides polynomial time solutions to all problems in NP.
I don't guarantee the results, but it should be good enough <wink>.

Jeremy




From skip at mojam.com  Thu Nov 11 17:30:20 1999
From: skip at mojam.com (Skip Montanaro)
Date: Thu, 11 Nov 1999 16:30:20 -0600 (CST)
Subject: Beginning programming with Python
In-Reply-To: <sng0ychgwt.fsf@motorola.com>
References: <sng0ychgwt.fsf@motorola.com>
Message-ID: <14379.17276.267502.673474@dolphin.mojam.com>

    Aaron> How quickly can I expect to pick up Python?

You should be comfortable with the syntactic elements of Python in a few
minutes (modulo the indentation-based block structure that tends to
initially derail people with a Perl/C/Pascal background) and presuming you
are familiar with object-oriented programming, comfortable with the language
as a whole within a few hours.  Unlike Perl, where much of the power is in
the language itself (e.g., its regex machinery), much of Python's
functionality lies in modules that you import when you need them (see
http://www.python.org/doc/lib/modindex.html, for example).  Getting
comfortable with a large fraction of that will take awhile.  There are
probably a half-dozen modules that you'll need to master fairly quickly and
pass on to your students, however: sys, string and os come to mind as well
as several of the builtin functions (available also via the module
__builtin__ and referred to in the library manual that way), and perhaps re
and time.

    Aaron> Secondly, if I choose to use Python for this, I would like to
    Aaron> supply each student with a reference book on Python.

>From what I've seen on the list, Learning Python sounds like a more
appropriate book for first-time programmers, though I've personally not seen
it.  Also, I would think that others, like me, who learned Python well
before any books were available and already knew other languages might be
ill-equipped to make unbiased recommendations in this area.  I have PP, and
while it is an excellently written book, I'd grown so used to using the
online docs, scanning the newsgroup archives and browsing source code that I
didn't find it added much to my existing sources of information, so I've
rarely used it.  You might want to scan through the reviews of both books at
Amazon.com to see what others have to say.

Finally, you might want to check out Guido van Rossum's Computer Programming
for Everybody (CP4E) project at

    http://www.python.org/doc/essays/cp4e.html

Skip Montanaro | http://www.mojam.com/
skip at mojam.com | http://www.musi-cal.com/
847-971-7098   | Python: Programming the way Guido indented...



From emuller at painewebber.com  Thu Nov 11 19:33:32 1999
From: emuller at painewebber.com (emuller at painewebber.com)
Date: Fri, 12 Nov 1999 00:33:32 GMT
Subject: Win32: CreateProcess wsetup program problem
References: <80fkki$kmg$1@nnrp1.deja.com>
Message-ID: <80fn8s$mls$1@nnrp1.deja.com>

I've also discoverd that winbatch's RunWait Function does not exhibit
this behavior.
In article <80fkki$kmg$1 at nnrp1.deja.com>,
  emuller at painewebber.com wrote:
> I wrote a wrapper to CreateProcess that runs a program with args and
> waits until it ends. It works great for most programs (ie..calc,
> notepad, net, etc, etc, etc), but when I use it to run a setup.exe
> program (NAV 5.0 to be exact), the setup program starts and an exit
> value of 0 is returned to me, but the setup program just started. I
> figure that the setup program is doing something, then spawning
another
> process to do the install (a _SETUP.EXE..something or other) and
> detaching itself from that process, letting it just run on....
>
> Is there anyone out there who knows how I can pause until that other
> program (or whatever happens) is done?
>
> I am using the following code in a module called win32helper.py
> (BTW: If anyone would like to use this code, go ahead. I hope to write
> a bunch of helper functions and put them here, then post them someday)
>
> import win32api
> import win32security
> import win32process
> import pywintypes
> import win32con
> import win32event
>
> def RunWait(cmd, args=""):
>         pHandle, tHandle, pId, tId = win32process.CreateProcess
(cmd," "
> + args,None,None,0,0,None,None,win32process.STARTUPINFO())
>         win32event.WaitForSingleObject(pHandle,win32event.INFINITE)
>         exitCode = win32process.GetExitCodeProcess(pHandle)
>         win32api.CloseHandle(pHandle)
>         win32api.CloseHandle(tHandle)
>         return exitCode
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
>


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



From skaller at maxtal.com.au  Tue Nov 30 08:52:05 1999
From: skaller at maxtal.com.au (skaller)
Date: Wed, 01 Dec 1999 00:52:05 +1100
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <81bmmi$279$1@nnrp1.deja.com> <3841EE77.6B5D6AC9@wjk.mv.com> <FLxz96.IEI@world.std.com>
Message-ID: <3843D685.C8C13355@maxtal.com.au>

Will Ware wrote:
> 
> William J. King (wjk at wjk.mv.com) wrote:
> : 4. conversion of integer to a string ( which may exist but havn't found
> : yet)
> 
> Try this:
>         str(123)
> or:
>         hex(123)
> 
> : 2. would like something like  'increment'/'decrement' operators
> :         either in form:
> :                     "x++"  or "incr x" vice "x = x + 1 " but its ok
> 
> I find myself grumbling about having to type "x = x + 1". The really
> clean thing to do, given that integers are objects, would be to define
> increment and decrement methods, so you'd type something like "i.incr()".

	Integers are NOT objects in Python, they're values.
Same for strings and tuples. We can pretend they're objects,
and say they're 'immutable', but the truth is that they're
just plain values.

	My new Viper interpreter may support 'mutable'
integers and strings as follows: you write:

	x := 1
	x++
	x+=1

This notation actually creates a reference (pointer) to the value,
so x is a 'reference to an integer', and x++ actually means:

	x := x + 1

To see the difference consider:

	a = 1
	x := a
	x++
	print a,x

and you will get 1,2. The integer value a is bound to is still
immutable and cannot be changed. x++, on the other hand,
does not increment what x refers to, rather it replaces what
x refers to by the old value plus 1. 

References in 'rvalue' contexts convert automatically to the
designated value, but when used on the LHS the distinction
between a reference to a value, and a value is preserved.
[References can be used with any type, but they're not
quite as useful with tuples (since there is a mutable list
already) or with class instance objects (since they're
already mutable).

Comments on this idea appreciated.

-- 
John Skaller, mailto:skaller at maxtal.com.au
10/1 Toxteth Rd Glebe NSW 2037 Australia
homepage: http://www.maxtal.com.au/~skaller
voice: 61-2-9660-0850



From AKiss at GEOMETRIA.hu  Tue Nov 16 07:57:10 1999
From: AKiss at GEOMETRIA.hu (Kiss, Arpad)
Date: Tue, 16 Nov 1999 13:57:10 +0100
Subject: Alright, I give up...chalk this one up to experience...
Message-ID: <01BF303A.7AFA1770@exchange1.geometria.hu>

Hi,

Maybe date is waiting for your new date setting, but why don't you use 
time instead of it:

import time
print time.ctime(time.time())
print time.localtime(time.time())

Arpad Kiss

Endgamer <endgamer71 at hotmail.com> wrote in message 
news:<383145c0.4445469 at 212.225.36.99>...
> Alright, I finally give up. I've been bashing at this bloody idea all
> week...maybe I'm incredibly stupid (probably), but I can't see why,
> when I run this as a script from Windows 95, it doesn't return the
> current date setting on my machine. Could somebody put me out of my
> misery? I promise I'll stop bothering you <g>.
> 
> -------------------------------------------------------
> 
> # Date setting program (chunk 1)
> import os
> 
> olddate=os.popen("date",'r')
> 
> olddate=readdate
> 
> print olddate
> 
> -------------------------------------------------------
> 
> 
> Endgamer
> Spider Warrior




From I.Clarke at strs.co.uk  Tue Nov 30 09:55:20 1999
From: I.Clarke at strs.co.uk (Ian Clarke)
Date: Tue, 30 Nov 1999 14:55:20 +0000
Subject: Python Make
Message-ID: <3843E558.8AC91A97@strs.co.uk>

Hi, just a quick idea I had that I wanted to share with the newsgroup. 
Has anyone considered using Python as a replacement for the "make"
utility?  A module could be provided which did several small things like
testing modification times of a list of source files against a
destination file to see whether an update was required.

I am prompted to suggest this after spending yet another half-day trying
to debug a Makefile.

What do people think?

Ian.

--
Ian Clarke                        Email: I.Clarke at strs.co.uk
Homepage: http://www.gnu.demon.co.uk/
Also see: http://freenet.on.openprojects.net/
"All we see and all we seem is but a dream within a dream"



From wheinemanNOwhSPAM at uconect.net.invalid  Sat Nov 13 01:16:15 1999
From: wheinemanNOwhSPAM at uconect.net.invalid (wheineman)
Date: Fri, 12 Nov 1999 22:16:15 -0800
Subject: Win32 API Communication Functions Module
Message-ID: <09920fb9.076b8f52@usw-ex0102-013.remarq.com>

I have Swigged the Win32 API Communication Functions and data
structures into a module. I've tested it on Windows 9x and NT4.0 with
Pythonwin 127. An example console terminal application is provided. The
module can be found at
http://www.uconect.net/~wheineman/python/win32comm.zip. I hope this is
useful to everyone looking for advanced serial port capabilities on
Win32 platforms.




* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!




From phd at phd.russ.ru  Sat Nov  6 06:34:44 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Sat, 6 Nov 1999 11:34:44 +0000 (GMT)
Subject: smtplib.py how to ?
In-Reply-To: <38237da0.3456244@news.force9.net>
Message-ID: <Pine.LNX.4.20.9911061132270.18974-100000@fep132.fep.ru>

On Sat, 6 Nov 1999, Colm Rafferty wrote:
> import smtplib
> s=smtplib.SMTP("localhost")
> tolist=["cgr at formulae.co.uk","c.rafferty at virgin.net","cgr at pangea.co.uk"]
> msg = '''
> From: test at my.org
> Subject: testin'...
> 
> This is a first test '''
> 
> s.sendmail("me at mydomain.co.uk",tolist,msg)
> s.quit()

   You've inserted empty line, so sendmail recognized it as "end of
headers". Do this way:

      vvvvv - no empty line here
msg = '''From: test at my.org
Subject: testin'...

This is a first test '''

or this:

      vvvv
msg = '''\
From: test at my.org
Subject: testin'...

This is a first test '''

Oleg.
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From mlh at vier.idi.ntnu.no  Fri Nov  5 09:29:11 1999
From: mlh at vier.idi.ntnu.no (Magnus L. Hetland)
Date: 05 Nov 1999 15:29:11 +0100
Subject: Drawing (directed) graphs
References: <000a01bf2792$b83ff150$3acbd9c2@peridot.optichrome.com> <y0jeme5owc9.fsf@vier.idi.ntnu.no> <14370.59564.979979.285996@gargle.gargle.HOWL>
Message-ID: <y0jbt99ovuw.fsf@vier.idi.ntnu.no>

Charles G Waldman <cgw at fnal.gov> writes:

> Magnus L. Hetland writes:
[...]
>  > Even on Minix you could easily generate jpegs, or postscript files or
>  > whatever with Piddle. If you can't run an X server or something, that
>  > only means that you can't see it... Or maybe you have a printer?
> 
> Or for maximum portatbility, you could use the "aa" ascii-art library
> to display the images on an ASCII terminal.... it's a C library, but
> it would be trivial to SWIG up a set of Python bindings, if this
> hasn't been done yet.

Excellennt suggestion :)

A nice backend for PIDDLE... Any takers?

--

  Magnus          Echelon jamming noise:
  Lie             FBI CIA NSA Handgun Assault Bomb Drug Terrorism
  Hetland         Special Forces Delta Force AK47 Hillary Clinton 



From wlfraed at ix.netcom.com  Fri Nov 26 20:54:11 1999
From: wlfraed at ix.netcom.com (Dennis Lee Bieber)
Date: Fri, 26 Nov 1999 17:54:11 -0800
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <al8yabo9fm3.fsf@sirppi.helsinki.fi> <86so1v4jlj.fsf@g.local> <al8hfiaan8p.fsf@sirppi.helsinki.fi> <81ksj6$lif$1@nntp9.atl.mindspring.net>
Message-ID: <aTU=OBhLWqlfXlzpQIGnHu1qIHZm@4ax.com>

On 26 Nov 1999 02:51:18 GMT, aahz at netcom.com (Aahz Maruch) declaimed the
following in comp.lang.python:

> In article <al8hfiaan8p.fsf at sirppi.helsinki.fi>,
> Markus Stenberg  <mstenber at cc.Helsinki.FI> wrote:
> >
> >From trivial example - do you consider (+ 2 (* 3 4)) more readable than 2 +
> >3 * 4 most languages (including Dylan) allow me to do?
> 
> I think that should be written as "3 4 * 2 +".  ;-)

	Taking it in the order the old HP manuals (and my old compiler
class) taught...

	2 3 4 * +
--
 > ============================================================== <
 >   wlfraed at ix.netcom.com  | Wulfraed  Dennis Lee Bieber  KD6MOG <
 >      wulfraed at dm.net     |       Bestiaria Support Staff       <
 > ============================================================== <
 >        Bestiaria Home Page: http://www.beastie.dm.net/         <
 >            Home Page: http://www.dm.net/~wulfraed/             <



From ionel at psy.uva.nl  Thu Nov 18 08:41:37 1999
From: ionel at psy.uva.nl (Ionel Simionescu)
Date: Thu, 18 Nov 1999 14:41:37 +0100
Subject: Comparing Python to (free) Math languages/environments
References: <19991118014716.13748.00000334@ng-bj1.aol.com>
Message-ID: <810vjh$7et@mail.psy.uva.nl>

Dparsavand <dparsavand at aol.com> wrote in message
news:19991118014716.13748.00000334 at ng-bj1.aol.com...
| I've browsed some interesting links from www.python.org and I have played
just
| a little with Python and Scilab.  I have used Matlab a fair amount in
school
| and find Octave to be a reasonable replacement.  I'm curious if anyone has
used
| any free Math scripting language or Math environment (Rlab, Scilab,
Octave,
| others?) as well as Python and might care to compare them.
|
| I tried to compare Octave to Python in terms of speed by producing some
random
| vectors and taking FFT's and Python seemed to by about twice as fast.
However,
| I'm not too keen on the syntax of Python for some things compared to
Octave
| (e.g. A*B vs. matrixmultiply(A,B)).  I read that one frustration with
Matlab
| (and I imagine Octave as well) is the limit to two-dimensional arrays,
which
| has bugged me in the past.
|
| I also read the CP4E proposal, and  I can definitely relate to the part
about
| how nice it would be to be able to modify the code in my GPS unit (mine
drives
| me crazy).  Python sounds great for these type of things, but does it
really
| make the best Math language?
|

Hi,

Python is a very nice language, to the extent it may be addicting.

However, the suitability of a certain language for a certain task, is also
determined by the availability of specialized libraries.

The support for (possibly multi-dimensional, heterogeneous) array
manipulation, some numerically intensive jobs and scientific visualization
is very good in the Python library.
(See, the relevant SIGs (Special Interest Groups) for pointers.)

However, this support is pretty generic.
If you have very specialized problems in fields such as DSP, ODE's, Kalman
filtering, time series, advanced statistics, then you better use one of the
field experts:
MATLAB/Mathematica/O-Matrix/Scilab/Octave/Gauss/Mathcad/etc.

"Better" does not mean that you'll be very happy coding, but you will get
*safer* results *faster*. You shall also find a lot more support in those
communities to your job specific problems.

If your problem can not benefit much from existing solutions, then Welcome!,
Python will provide you the means to solve your problem.

As you mentioned performance issues, allow me to notice that, indeed, you
can code in Python such as to beat MATLAB, but you will still lag behind
C/C++/Fortran. If speed is essential you should be prepared to consider your
Python code a prototype and translate it to a lower-level, optimized
implementation.

Very often, the < design&test with Python -implement with C/C++ > approach
it's still the fastest way to come with a solution.


ionel


PS
MATLAB currently (5.#) supports multi-dimensional, heterogeneous arrays,
data structures and even some form of OOP (actually, just function
overloading).






From idt at ornl.gov  Mon Nov  1 13:48:07 1999
From: idt at ornl.gov (Russell W. Schmidt)
Date: Mon, 01 Nov 1999 18:48:07 GMT
Subject: Contribution: nice floating point number formatting
References: <381a4316$0$209@nntp1.ba.best.com> <19991030063317.A16991@better.net> <7vhrfm$glv$1@vvs.superst.iae.nl> <19991101124124.A903@Ridcully.home>
Message-ID: <7vkn97$8ts$1@sws1.ctd.ornl.gov>

First off - thanks to the original contributor for a nice addition to my 
toolbox.

In article <19991101124124.A903 at Ridcully.home>, "Malcolm Tredinnick" 
<malcolmt at smart.net.au> wrote:
>On Sun, Oct 31, 1999 at 05:41:26PM +0100, Carel Fellinger wrote:
>> And for the spaces in the decimal part, are you shure that is conventional?
>> I can't remember having seen this *ever* in my life. At least not in my life
>> as a programmer.
>
>There is a convention in some situations/places where the comma (or, in
>Europe, the dot) is dropped entirely and a thin space is used (if you are
>typesetting -- something smaller than a normal inter-word space).
>
>You will notice I have couched that last paragraph somewhat cautiously: my
>recollection is that this situation (using spaces) is meant to be standard,
>these days but since everybody is used to using commas, etc, they still
>appear quite frequently. IMVHO, the spaces every three digits look nicer
>than throwing a bunch of punctuation marks at the number. :-)
>
>(and if that doesn't start some sort of religious war, I don't know what
>will.) :-)
>

The spaces every three digits, on both sides of the decimal point, are indeed 
part of an international standard, called the International System of Units 
(the Metric System to most Americans).  You can fnd a US reference to it on 
the NIST (National Institute of Standards and Technology) web site. It is 
specifically called out in NIST Special Publication 811 (1995 Edition) "Guide 
for the Use of the International System of Units (SI)".  See section 10.5.3, 
Grouping Digits. The authors note that chapter 10 is adapted in part from ISO 
31-0 and ISO 31-11. 

I will not type in the whole section (pull it from NIST if you care), but the 
short summary in their "Check List for Reviewing Manuscripts" says:

  "The digits of numerical values having more than four digits on either 
    side of the decimal marker are separated into groups of three using a
    thin, fixed space counting from both the left and right of the decimal 
    marker. For example, 15 739.012 53 is highly preferred to 15739.01253.
    Commas are not to be used to separate digits into groups of three."


--
Russ Schmidt         (idt at ornl.gov)
Lockheed Martin Energy Systems, Inc.
Oak Ridge, Tennessee     37831-2009



From gerrit.holl at pobox.com  Wed Nov 10 11:12:56 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Wed, 10 Nov 1999 17:12:56 +0100
Subject: Intelligent strptime?
Message-ID: <19991110171256.A5196@optiplex.palga.uucp>

Hello,

Does there exist an intelligent strptime? "26/11/99", for example, can _only_
mean the 26th November of this year, and the chance that "14:11:12" means
14 november 1912 is very small. Is there a module/function that guesses the
format? There isn't doubt very often, only for 12/11/99, US is 11 december,
Europe is 12 November (isn't it?).

regards,
Gerrit.



From greg.ewing at compaq.com  Fri Nov 19 04:15:45 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Fri, 19 Nov 1999 22:15:45 +1300
Subject: Type/Class Unification in Ruby
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp> <7vjtrc$hmm$1@newshost.accu.uu.nl> <38204614.2EAC6051@compaq.com> <5ld7trtak5.fsf@eric.cnri.reston.va.us> <3826D63C.2731BB1F@compaq.com> <38328948.9E35556F@compaq.com> <m3d7t9v8lr.fsf@qiao.localnet>
Message-ID: <38351541.FD1EBDEF@compaq.com>

Clemens Hintze wrote:
> 
> But it is worth mentioning, that Ruby features
> single inheritance *by*purpose*. Not because of the implementation
> details!

I wasn't making a value judgement, only pointing out that
Ruby's solution is as simple as it is because of this
restriction. 

> If it would implement multi-inheritance, the class structure could
> e.g. use an array of pointer to all such C structs, not only one
> pointer, as implemented now!

Sure, Python could do something like that too. But it would
get very hairy and complicated and probably slow everything
down a lot, casting doubt on whether it's worth the effort.

A middle ground might be to allow inheriting from any number
of user-defined classes, but at most one built-in type.

> The second (preferred) way is, to build instances of class DATA
> wrapping the C struct you have allocated for your extension.

Yes, I know, but what I was saying still applies in that case.
All the methods of the extension class expect the DATA object
to be pointing to their own particular kind of struct, and
because of the inheritance structure, this is always true --
so they don't need to bother checking.

> Deriving a Ruby class coded in C from another one also coded in C,
> seems to be the only right tricky part of that language, IMHO. :-)

I think it will work as long as the derived class makes sure
that its C structure is an exension of the superclass's C
structure.

>     Greg> [global instance var dict] might cause ref counting
>     Greg> problems.
> 
> Once again, thanks for Ruby's true garbage collection! :-)

Garbage collection alone doesn't actually solve the problem;
the references from the global dict need to be treated as
weak references somehow under either scheme. I haven't
looked, but presumably Ruby must be ignoring the global dict
refs when marking, and cleaning them up whenever a builtin
or extension object is deallocated. Python could be made to
do the equivalent.

> The big exception,
> AFAIK, was class Fixnum. It never had any instanciation method, as it
> doesn't wrap any struct.

Yes, that's the one class you really can't instantiate a
subclass of, since its class is totally implicit.

But it's not clear whether it's any use having something
like a UserFlonum, either, even though technically it could
be done. Unless you override every method, any arithmetic
on it would return an ordinary Flonum. And if you are going
to override every method, it might as well be a totally
new class!

> Often it is better to use the module
> mechanism to add methods to class Array or some of its instances
> during runtime.

You mean you can add methods to an *existing* class that
way? Not sure if I like that idea...

Greg



From kuntsi at cc.helsinki.fi.spam.no  Tue Nov 16 05:10:07 1999
From: kuntsi at cc.helsinki.fi.spam.no (Antti Kuntsi)
Date: 16 Nov 1999 10:10:07 GMT
Subject: Memory not being released?
References: <1269432569-4253767@hypernet.com>
Message-ID: <slrn4832bbv.eqh.kuntsi@sirppi.helsinki.fi>

Gordon McMillan <gmcm at hypernet.com> sepusti:
> Steve Tregidgo writes:
> > I have a process (goes up to 250+ MB, cue lots of virtual memory
> > usage and hence slowdown), which busies the server so much that
> > it just can't bring itself to do other things -- in other words
> > it's severely impractical.
> That's a pretty good sign that you have circular references that 
> are keeping objects alive. Check out Cyclops from 
> ftp://ftp.python.org/pub/python/contrib/System/.

I had similar problem just a while ago, and Cyclops helped a lot. But it
didn't solve the whole problem, as it was not python's fault, but a
Linux library glibc was functioning improperly. If you are running RedHat 6.0
and/or have glibc version 2.1.1, you should upgrade asap. Upgrading to
2.1.2 the bloating dissappeared and everything has worked nicely ever
since.

Antti Kuntsi

--
______________________________________________________________
|O| ----------------------------------------------------------
| | |Twisted mind? No, just bent in several strategic places.|
`\| |___________________________________mickut at iki.fi/mickut/|



From edwardm_painewebber at my-deja.com  Fri Nov 12 10:01:32 1999
From: edwardm_painewebber at my-deja.com (edwardm_painewebber at my-deja.com)
Date: Fri, 12 Nov 1999 15:01:32 GMT
Subject: .dll and .exe file version informaton
References: <816010E2456BD111A48700805FBBE2EEFDF793@ex-quebec-u1.baan.com>
Message-ID: <80ha4a$pmg$1@nnrp1.deja.com>

It looks like they are not available yet....
In article <816010E2456BD111A48700805FBBE2EEFDF793 at ex-quebec-
u1.baan.com>,
  Gaetan Corneau <Gaetan_Corneau at baan.com> wrote:
> > go-ahead-just-*try*-to-write-a-program-to-do-that<wink>-ly y'rs  -
tim
>
> If you really want to do it in a program, take a look at the following
> functions in the help:
> 	GetFileVersionInfo
> 	VerQueryValue
> 	QueryField
>
> You wil find all this stuff crappy, but it works.
> ______________________________________________________
>    Gaetan Corneau
>    Software Developer (Quality Assurance Team)
>    BaaN  Supply Chain Solutions
>    E-mail: Gaetan_Corneau at baan.com
>    Compuserve: Gaetan_Corneau at compuserve.com
>    ICQ Number: 7395494
>    Tel: (418) 266-8252
> ______________________________________________________
> "Profanity is the one language all programmers know best"
>
>


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



From pinard at iro.umontreal.ca  Tue Nov  2 09:27:46 1999
From: pinard at iro.umontreal.ca (=?ISO-8859-1?Q?Fran=E7ois_Pinard?=)
Date: 02 Nov 1999 09:27:46 -0500
Subject: Printing strings
In-Reply-To: Matthew Hirsch's message of "Tue, 02 Nov 1999 10:18:59 -0500"
References: <381F00E1.721BB47F@cornell.edu>
Message-ID: <oqso2pnf31.fsf@titan.progiciels-bpi.ca>

Matthew Hirsch <meh9 at cornell.edu> writes:

>    a=4.0
>    b=2.0
>    print '2(%3.1f)' % a,b
> where I really mean
>    print '%3.1f %3.1f' % a,b
> but it's printing
>    2(4.0) 2.0
> Is there a way of doing this so I don't have to write out %3.1f,
> multiple times?

Maybe you could play around this idea?

>>> a, b = 4.0, 2.0
>>> print '%4.1f' * 2 % (a, b)
 4.0 2.0

I changed `%3.1f' into `%4.1f' because `4.02.0' would not look nice.)

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





From phd at phd.russ.ru  Tue Nov 16 11:45:38 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Tue, 16 Nov 1999 16:45:38 +0000 (GMT)
Subject: Anyway to use file masks with os.listdir()?
In-Reply-To: <Pine.SO4.4.02.9911160925560.22506-100000@ub.d.umn.edu>
Message-ID: <Pine.LNX.4.20.9911161645190.2604-100000@fep132.fep.ru>

On Tue, 16 Nov 1999, Aaron J Reichow wrote:
> Is there any way to pass a filemask to os.listdir()?  For example, to list
> only files with the extension .jpg.  Any help would be appreciated! :)

   Module fnmatch.py is your friend.

Oleg.
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From matz at netlab.co.jp  Sun Nov  7 21:26:30 1999
From: matz at netlab.co.jp (Yukihiro Matsumoto)
Date: 08 Nov 1999 11:26:30 +0900
Subject: A Comparison of Python and Ruby Extension Interfaces (Re: Comparison   between Python and "Ruby")
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp>
 <7vjtrc$hmm$1@newshost.accu.uu.nl> <38204614.2EAC6051@compaq.com>
 <021119992327527678%pereira@research.att.com>
 <38215CE7.70C3D560@compaq.com> <SzUjOCp3Jbh1lZd9WdVDdXeTiSia@4ax.com>
Message-ID: <87vh7du3ah.fsf@ev.netlab.co.jp>

Toby Dickenson <htrd90 at zepler.org> writes:

|>It uses setjmp/longjmp. Presumably these work correctly on
|>any system where they are implemented. Porting Ruby to a
|>system where they're not implemented could be difficult.
|>
|
|Presumably this makes it difficult for Ruby extensions to use C++?

I don't think so.  Some extensions for Ruby are written in C++
(e.g. MySQL interface) without problems.

							matz.



From philh at vision25.demon.co.uk  Mon Nov 15 20:57:23 1999
From: philh at vision25.demon.co.uk (Phil Hunt)
Date: Tue, 16 Nov 99 01:57:23 GMT
Subject: __str__ vs. __repr__
References: <000001bf275c$02370d20$1d2d153f@tim> <7vvni7$19q0$1@nntp6.u.washington.edu> <382BEC79.EC65913D@maxtal.com.au> <382C6430.8DC07E3E@lemburg.com> <01fd01bf2dce$d2c03cf0$f29b12c2@secret.pythonware.com> <942499879snz@vision25.demon.co.uk> <m3n1shp2uu.fsf@atrus.jesus.cam.ac.uk>
Message-ID: <942717443snz@vision25.demon.co.uk>

In article <m3n1shp2uu.fsf at atrus.jesus.cam.ac.uk>
           mwh21 at cam.ac.uk "Michael Hudson" writes:
> philh at vision25.demon.co.uk (Phil Hunt) writes:
> 
> > In article <01fd01bf2dce$d2c03cf0$f29b12c2 at secret.pythonware.com>
> >            fredrik at pythonware.com "Fredrik Lundh" writes:
> > > 2) what's the reason byte code cannot be used to recover
> > > the original source code
> > 
> > Perhaps because two different (but obviously functionally
> > identical and very similar) source codes could produce the same
> > byte code.
> 
> I doubt it actually. Python's bytecode is really pretty high
> level. Leastaways, I can't think of any examples.

How about:

   x = a + b + c
   x = (a + b) + c
   x=a+b+c

I'd guess these all produce the same bytecode.

I'd also not be surprised if you could consistently change the
name of a local variable within a function without changing bytecodes,
i.e. Python internally doesn't know the names of local variables.

> It's one of those things I might get round to writing one of these
> days. I don't really know how hard it would be. Not especially, I
> suspect.

You might want to have a look at how Java bytecode decompilers
work.

-- 
Phil Hunt - - - - - - - - -  philh at vision25.demon.co.uk
Moore's Law: processor speed doubles every 18 months
Gates' Law: software speed halves every 18 months




From a.eyre at optichrome.com  Thu Nov 25 09:05:02 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Thu, 25 Nov 1999 14:05:02 -0000
Subject: decompress a bunch of .gz files
In-Reply-To: <81jc8i$ilp$1@nnrp1.deja.com>
Message-ID: <000501bf374e$116bafb0$3acbd9c2@peridot.optichrome.com>

> I'm searching for a way to decompress all .gz
> files in a directory automatically.

Unix:

#!/bin/sh
for file in *.gz; do gunzip $file; done


Dos:

@echo off
for %%file in (*.gz) do gunzip %%file

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From edwardam at home.com  Tue Nov 30 00:10:09 1999
From: edwardam at home.com (Edward Muller)
Date: Tue, 30 Nov 1999 05:10:09 GMT
Subject: SMTP
References: <19991126095821.87186.qmail@hotmail.com> <38401AD2.31396C84@wjk.mv.com>
Message-ID: <38435C2D.435B3986@home.com>

"William J. King" wrote:

> Was just looking at an interesting program called Rebol which does some
> interesting things with mail/ftp etc. http://www.rebol.com - handles
> various protocols etc. A scripting communications/messaging language..
> hope I got that correct.. Sassenfrath of Amiga fame is source.
>
> This might be yet another simple solution for  SMTP area.

Have ever tried to understand a long Rebol Script? I've tried. I thought
it was really great at first, but then got a BIG headache. Python is much
easier to read




From Gaetan_Corneau at baan.com  Tue Nov 16 17:12:54 1999
From: Gaetan_Corneau at baan.com (Gaetan Corneau)
Date: Tue, 16 Nov 1999 17:12:54 -0500
Subject: Time object (COM)
Message-ID: <816010E2456BD111A48700805FBBE2EEFDF7D7@ex-quebec-u1.baan.com>

Hi,

Sorry to ask this again, I can't find the e-mails I received when I first
asked :(

I have a COM object with a property that looks like this when I print it:
<time object at c5d430>

What is it exactly? Where is it documented? 

Thanks,
______________________________________________________
   Gaetan Corneau
   Software Developer (Quality Assurance Team)
   BaaN  Supply Chain Solutions  
   E-mail: Gaetan_Corneau at baan.com        
   Compuserve: Gaetan_Corneau at compuserve.com  
   ICQ Number: 7395494             
   Tel: (418) 266-8252          
______________________________________________________
"Profanity is the one language all programmers know best"




From Matija.Exel at lag.ensieg.inpg.fr  Tue Nov  9 06:31:57 1999
From: Matija.Exel at lag.ensieg.inpg.fr (Matija Exel)
Date: Tue, 09 Nov 1999 12:31:57 +0100
Subject: build failure for SunOS 5.5
References: <38269571.14471A7D@lag.ensieg.inpg.fr> <38278EAF.1C6055E2@python.net>
Message-ID: <3828062D.B2547E7F@lag.ensieg.inpg.fr>



> Those sound like pretty severe messages in the configure.log.  Did
> ./configure run to completion without error?

yes, no errror on ./configure

> Can you tell which
> portion of the configure script generates the error messages?

well, here is the complete config.log:
__________________________
This file contains any messages produced by compilers while
running configure, to aid debugging if configure makes a mistake.

configure:608: checking MACHDEP
configure:653: checking CCC
configure:669: checking for --without-gcc
configure:746: checking for gcc
configure:859: checking whether the C compiler (cc  ) works
configure:875: cc -o conftest    conftest.c  1>&5
configure:901: checking whether the C compiler (cc  ) is a cross-compiler
configure:906: checking whether we are using GNU C
configure:934: checking whether cc accepts -g
configure:997: checking LINKCC
configure:1014: checking LDLIBRARY
configure:1043: checking for ranlib
configure:1076: checking for ar
configure:1150: checking how to run the C preprocessor
configure:1230: checking for AIX
configure:1255: checking for minix/config.h
configure:1304: checking whether cc accepts -OPT:Olimit=0
configure:1338: checking whether cc accepts -Olimit 1500
configure:1375: checking for C preprocessor type
configure:1388: cc -c -g  conftest.c 1>&5
"configure", line 1381: invalid source character: '#'
"configure", line 1381: warning: improper pointer/integer combination: op
"="
"configure", line 1381: non-constant initializer: op "NAME"
"configure", line 1381: invalid source character: '#'
"configure", line 1381: warning: improper pointer/integer combination: op
"="
"configure", line 1381: non-constant initializer: op "NAME"
"configure", line 1381: syntax error before or at: "() -- "
"configure", line 1381: warning: syntax error:  empty declaration
configure: failed program was:
#line 1377 "configure"
#include "confdefs.h"

#define spam(name, doc) {#name, &name, #name "() -- " doc}
int foo;
struct {char *name; int *addr; char *doc;} desc = spam(foo, "something");

int main() {
;
; return 0; }
configure:1405: checking for ANSI C header files
configure:1516: checking for dlfcn.h
configure:1516: checking for fcntl.h
configure:1516: checking for limits.h
configure:1516: checking for locale.h
configure:1516: checking for ncurses.h
configure:1516: checking for pthread.h
configure:1516: checking for signal.h
configure:1516: checking for stdarg.h
configure:1516: checking for stddef.h
configure:1516: checking for stdlib.h
configure:1516: checking for thread.h
configure:1516: checking for unistd.h
configure:1516: checking for utime.h
configure:1516: checking for sys/audioio.h
configure:1516: checking for sys/file.h
configure:1516: checking for sys/lock.h
configure:1516: checking for sys/param.h
configure:1516: checking for sys/select.h
configure:1516: checking for sys/time.h
configure:1516: checking for sys/times.h
configure:1516: checking for sys/un.h
configure:1516: checking for sys/utsname.h
configure:1516: checking for sys/wait.h
configure:1557: checking for dirent.h that defines DIR
configure:1595: checking for opendir in -ldir
configure:1681: checking for clock_t in time.h
configure:1710: checking for mode_t
configure:1743: checking for off_t
configure:1776: checking for pid_t
configure:1809: checking return type of signal handlers
configure:1850: checking for size_t
configure:1883: checking for uid_t in sys/types.h
configure:1918: checking size of int
configure:1957: checking size of long
configure:1996: checking size of void *
configure:2036: checking for long long support
configure:2046: cc -c -g  conftest.c 1>&5
configure:2060: checking size of long long
configure:2102: checking size of off_t
configure:2144: checking whether to enable large file support
configure:2170: checking for --with-next-framework
configure:2187: checking for --with-dyld
configure:2213: checking SO
configure:2227: checking LDSHARED
configure:2267: checking CCSHARED
configure:2289: checking LINKFORSHARED
configure:2324: checking for dlopen in -ldl
configure:2371: checking for shl_load in -ldld
configure:2424: checking for t_open in -lnsl
configure:2464: checking for socket in -lsocket
configure:2504: checking for socket in -lnet
configure:2547: checking for --with-libs
configure:2561: checking for --with(out)-readline
configure:2576: checking for --with-dec-threads
configure:2592: checking for --with-threads
configure:2606: checking for --with-thread
configure:3113: checking for --with-sgi-dl
configure:3136: checking for --with-dl-dld
configure:3169: checking for alarm
configure:3169: checking for chown
configure:3169: checking for clock
configure:3169: checking for dlopen
configure:3169: checking for execv
configure:3169: checking for flock
configure:3169: checking for fork
configure:3169: checking for fsync
configure:3169: checking for fdatasync
configure:3169: checking for ftime
configure:3169: checking for ftruncate
configure:3169: checking for getpeername
configure:3169: checking for getpgrp
configure:3169: checking for getpid
configure:3169: checking for getpwent
configure:3169: checking for gettimeofday
configure:3169: checking for getwd
configure:3169: checking for kill
configure:3169: checking for link
configure:3169: checking for lstat
configure:3169: checking for mkfifo
configure:3169: checking for mktime
configure:3169: checking for nice
configure:3169: checking for pause
configure:3169: checking for plock
configure:3169: checking for pthread_init
configure:3169: checking for putenv
configure:3169: checking for readlink
configure:3169: checking for select
configure:3169: checking for setgid
configure:3169: checking for setlocale
configure:3169: checking for setuid
configure:3169: checking for setsid
configure:3169: checking for setpgid
configure:3169: checking for setpgrp
configure:3169: checking for setvbuf
configure:3169: checking for sigaction
configure:3169: checking for siginterrupt
configure:3169: checking for sigrelse
configure:3169: checking for strftime
configure:3169: checking for strptime
configure:3169: checking for symlink
configure:3169: checking for tcgetpgrp
configure:3169: checking for tcsetpgrp
configure:3169: checking for timegm
configure:3169: checking for times
configure:3169: checking for truncate
configure:3169: checking for uname
configure:3169: checking for waitpid
configure:3226: checking for fseek64
configure:3226: checking for fseeko
configure:3226: checking for fstatvfs
configure:3226: checking for ftell64
configure:3226: checking for ftello
configure:3226: checking for statvfs
configure:3282: checking for dup2
configure:3282: checking for getcwd
configure:3282: checking for strdup
configure:3282: checking for strerror
configure:3282: checking for memmove
configure:3337: checking for getpgrp
configure:3387: cc -c -g  conftest.c 1>&5
configure:3403: checking for setpgrp
configure:3453: cc -c -g  conftest.c 1>&5
configure:3469: checking for gettimeofday
configure:3519: cc -c -g  conftest.c 1>&5
configure:3538: checking whether time.h and sys/time.h may both be
included
configure:3573: checking whether struct tm is in sys/time.h or time.h
configure:3607: checking for tm_zone in struct tm
configure:3640: checking for tzname
configure:3678: checking for time.h that defines altzone
configure:3712: checking whether sys/select.h and sys/time.h may both be
included
configure:3725: cc -c -g  conftest.c 1>&5
configure:3741: checking whether char is unsigned
configure:3805: checking for working const
configure:3882: checking for working volatile
configure:3891: cc -c -g  conftest.c 1>&5
"configure", line 3887: warning: volatile is a keyword in ANSI C
"configure", line 3887: undefined symbol: volatile
"configure", line 3887: syntax error before or at: int
"configure", line 3887: undefined symbol: x
configure: failed program was:
#line 3884 "configure"
#include "confdefs.h"

int main() {
volatile int x; x = 0;
; return 0; }
configure:3908: checking for working signed char
configure:3917: cc -c -g  conftest.c 1>&5
"configure", line 3913: warning: signed is a keyword in ANSI C
"configure", line 3913: undefined symbol: signed
"configure", line 3913: syntax error before or at: char
configure: failed program was:
#line 3910 "configure"
#include "confdefs.h"

int main() {
signed char c;
; return 0; }
configure:3934: checking for prototypes
configure:3943: cc -c -g  conftest.c 1>&5
"configure", line 3940: warning: statement not reached
configure:3958: checking for variable length prototypes and stdarg.h
configure:3977: cc -c -g  conftest.c 1>&5
"configure", line 3965: va_start: argument mismatch
"configure", line 3974: warning: statement not reached
configure:3993: checking for bad exec* prototypes
configure:4002: cc -c -g  conftest.c 1>&5
configure:4019: checking for bad static forward
configure:4040: cc -o conftest -g   conftest.c -lsocket -lnsl -ldl  1>&5
configure:4059: checking whether va_list is an array
configure:4074: cc -c -g  conftest.c 1>&5
configure:4090: checking for gethostbyname_r
configure:4138: checking gethostbyname_r with 6 args
configure:4159: cc -c -g     conftest.c 1>&5
configure:4333: checking for __fpu_control in -lieee
configure:4382: checking for --with-fpectl
configure:4405: checking for --with-libm=STRING
configure:4426: checking for --with-libc=STRING
configure:4450: checking for hypot
configure:4505: checking for hypot
configure:4563: checking for genuine getopt
configure:4601: checking what malloc(0) returns
_______end config.log


--
-------------------------------------------------------------------
 M. Matija Exel
 E.N.S.I.E.G., Service  Informatique / L.A.G.
 BP.  46 Cedex,  38402 St.Martin d'Heres,  FRANCE
 Tel : (+33) 4 76 82 71 12          Fax:(+33) 4 76 82 63 88
 e-mail : Matija.Exel at lag.ensieg.inpg.fr
 web:  http://www-hadoc.ensieg.inpg.fr/~exel/





From free at money.com  Thu Nov 25 16:20:43 1999
From: free at money.com (free at money.com)
Date: Thu, 25 Nov 1999 15:20:43 -0600
Subject: Make money surfin the web!            -- 89191
Message-ID: <251199152043@money.com>

Everyone likes making easy money.  This place pays you 50 cents for every hour you browse the web!!!
Get even more if you refer people!  I just got a check for $36.00!!!

Sign up here:
http://www.alladvantage.com/go.asp?refid=GOB631








r2WYjzrKf-



From dgobbi at hickory.irus.rri.on.ca  Mon Nov  1 15:43:27 1999
From: dgobbi at hickory.irus.rri.on.ca (David Gobbi)
Date: 1 Nov 1999 20:43:27 GMT
Subject: __str__ vs. __repr__
References: <19991101121709.A134039715@vislab.epa.gov> <19991101202118.B29736@optiplex.palga.uucp>
Message-ID: <7vku1f$50r$1@panther.uwo.ca>

Gerrit Holl <gerrit.holl at pobox.com> wrote:
> Randall Hopper wrote:
>> 
>> 1) Is there a convention for what __str__ and __repr__ should return for
>>    classes?

> __str__ should return a string, __repr__ may also return, for example, a
> dictionairy.

The Python Language Reference Manual has a few lines on this subject,
specifically in Section 3.3.1.  

__repr__ (self) 
       Called by the repr() built-in function and by string conversions 
       (reverse quotes) to compute the "official" string representation 
       of an object. This should normally look like a valid Python
       expression that can be used to recreate an object with the same 
       value. By convention, objects which cannot be trivially converted 
       to strings which can be used to create a similar object produce a
       string of the form "<...some useful description...>". 

__str__ (self) 
       Called by the str() built-in function and by the print statement 
       to compute the "informal" string representation of an object. 
       This differs from __repr__() in that it does not have to be a valid
       Python expression: a more convenient or concise representation may 
       be used instead. 

In short:  no they are not the same, yes you can define them both.  If you
           don't define __str__(), then str(object) will default to
           repr(object)

  - David
>> 2) Or, whatever they return, should they return the same value?

> I don't know: I think it's wrong to use both.

> regards,
> Gerrit.
> -- 
> linuxgames.nl.linux.org	 All about games on Linux in Dutch (under construction).
> www.nl.linux.org	       The Dutch resource for Dutch Linux documentation.
> www.nl.linux.org/discoverb		    Learn foreign words and definitions.
> www.nl.linux.org/~gerrit/asperger	 I'm an Asperger, go here for more info.
> www.nl.linux.org/~gerrit			    About me and pictures of me.



-- 
--David Gobbi, MSc                    dgobbi at irus.rri.on.ca
  Advanced Imaging Research Group
  Robarts Research Institute, University of Western Ontario



From mhammond at skippinet.com.au  Tue Nov  9 22:15:25 1999
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 10 Nov 1999 03:15:25 GMT
Subject: MS Word COM error w/ Borders property
References: <38289ED8.9BA2645A@alliedsignal.com>
Message-ID: <hr5W3.117$Xo.955@news-server.bigpond.net.au>

john whittaker wrote in message <38289ED8.9BA2645A at alliedsignal.com>...
>I got the following error when I tried to use get at the bottom border
>from a borders collection:
>>>> borders = word.ActiveDocument.Tables[0].Borders
>>>>borders[ -3 ]

Try using borders.Item(-3)

Using the "-3" syntax creates an enumerator, in which the -ve index is
invalid.

Mark.






From dieter at handshake.de  Wed Nov 10 14:16:49 1999
From: dieter at handshake.de (Dieter Maurer)
Date: 10 Nov 1999 20:16:49 +0100
Subject: Popup menus in Tkinter
References: <199911081726.SAA25412@axil.hvision.nl> <80796f$f9h$1@nnrp1.deja.com>
Message-ID: <x7u2mu6tsu.fsf@lindm.dm>

dkaznadzey at my-deja.com writes on Mon, 08 Nov 1999 19:44:16 GMT:
> What would be a good source of information on tk commands
> available through Tkinter?
I use "tclhelp", a program with comes with "tclX".
It is a grafical documentation program for "tcl", "tclX"
and "tk". Alternatively, you can use the online tcl/tk
documentation (e.g. from scriptics).

These, of cause, document "tcl/tk" not "Tkinter".
However, the mapping from "tk" to "Tkinter" is quite straight forward:

* If a widget "w" supports a command "c" with positional parameters
  "p1", "p2", ... and option value pairs "o1", "v1", "o2", "v2", ...,
  then the corresponding "Tkinter" object "W" usually has a method "c"
  which can be called: W.c(p1,p2,..., o1=v1, o2=v2, ...)

  Examples:
    a text widget "t" supports the "insert" command with
    parameters "index", "chars" and optionally "taglist",
    "chars", "taglist", "chars" and so on.
    So one can call:

        t insert end Hello tag1 'Bye' {tag2 tag3}

    For the corresponding Tkinter object "T", this looks like:

        T.insert('end','Hello','tag1','Bye',['tag2','tag3'])


    the "configure" command of a widget may take several
    forms:

      a)        w configure
      b)        w configure -option
      c)        w configure -option1 value1 -option2 value2 ...

    In Tkinter, this looks like:

      a)        w.configure()
      b)        w.configure('option')
      c)        w.configure(option1=value1, option2=value2, ...)


* If a widget "w" has "subobjects" of type "s" which supports
  command "c", then the corresponding "Tkinter" method
  is "W.s_c".
  Positional parameters and keyword parameters (option,value pairs
  in tcl) are handled as shown above.

  Example:
    a text widget "t" has "tag" subobjects supporting e.g.
    the command "bind".

    The "Tkinter" method is "T.tag_bind".


* Tk control commands (e.g. "pack", "bind", "grid", wm, winfo, ...)
  are mapped to corresponding method calls of the first slave.

  Examples:
    a)          pack slave1 slave2 .... -option1=value1 -option2=value2 ...
    b)          grid slave -option1=value1 ...

    become in Tkinter:
    a)          slave1.pack(option1=value1, option2=value2, ...)
                slave2.pack(option1=value1, option2=value2, ...)
                ....
    b)          slave.grid(option1=value1, ...)

  A sub-commands "s" of a control command "c" is mapped to
  the method "s_c".

  Example:
      pack forget slave1 slave2 ...
    becomes:
      slave1.pack_forget()
      slave2.pack_forget()
      ....


There are few exceptions from the rules stated above, e.g.:
  * additional parameter "add" for the various forms of "bind"
    to tell Tkinter to add a function to the action list
    (or replace the action list).
  * special handling of variables (e.g. bound to buttons):
    they must be created as Tkinter objects while in
    tcl/tk creation is automatic.


- Dieter



From jam at quark.emich.edu  Sun Nov 21 11:48:27 1999
From: jam at quark.emich.edu (Jeff)
Date: Sun, 21 Nov 1999 11:48:27 -0500
Subject: Order of keyword arguments
In-Reply-To: <38380F92.D9817451@corrada.com>
References: <38379233.ADE65CDB@corrada.com> <19991121034245.A21255@quark.emich.edu> <38380F92.D9817451@corrada.com>
Message-ID: <19991121114827.A21697@quark.emich.edu>

On Sun, Nov 21, 1999 at 10:28:18AM -0500, Andres Corrada wrote:
> 
> I want to have my cake and eat it too! I thought this would be a way of
> keeping the flexibity of keyword arguments but also pass information
> along on how the calling function wants the items returned. For example,
> an HTML form could be called with a variable number of input tags and
> you want to render the HTML in the same order that they were used to
> call the function.
> 

hmmm.. well, you can't do it that way ;)

you can write your function such that the order in which the arguments are
processed is static (using kw.items() and a series of if..then statements in
some specific order) in order to get this done, but I don't see the
advantage to 'knowing' the order that the keywords were used by the calling
function.

I have used something like this:

def foo(**kw):
 for k,v in kw.items():
  if k == "bar":
   print "bar keyword value:",v
  elif k == "baz":
   print "baz keyword value:", v
 .
 .
 .


in code that was building arguments to pass to ipchains, and it seems to
work OK. does that help?

regards,
J
-- 
|| visit gfd <http://quark.emich.edu/>
|| psa member #293 <http://www.python.org/> 
|| New Image Systems & Services, Inc. <http://www.newimage.com/>



From Gaetan_Corneau at baan.com  Tue Nov 23 11:58:14 1999
From: Gaetan_Corneau at baan.com (Gaetan Corneau)
Date: Tue, 23 Nov 1999 11:58:14 -0500
Subject: Python and COM: a bug?
Message-ID: <816010E2456BD111A48700805FBBE2EEFDF81A@ex-quebec-u1.baan.com>

> If this is vital to your project, it might be worth running the source
> through a debugger.

I have debugged the C++ source (our code), and I'm sure it is OK (and the
problem never happens with lousy VBScript anyway).
I have traced in the Python code, but I could not pointexactly what's wrong,
but I think there *is* something wrong in there.
______________________________________________________
   Gaetan Corneau
   Software Developer (Quality Assurance Team)
   BaaN  Supply Chain Solutions  
   http://www.baan.com
   E-mail: Gaetan_Corneau at baan.com                  
   Tel: (418) 266-8252          
______________________________________________________
"Profanity is the one language all programmers know best"






From thomas.heller at ion-tof.com  Thu Nov  4 06:56:36 1999
From: thomas.heller at ion-tof.com (Thomas Heller)
Date: Thu, 4 Nov 1999 12:56:36 +0100
Subject: Embedding and threads
References: <001b01bf26ab$61c951f0$0501a8c0@bobcat>
Message-ID: <002901bf26bb$a601aa10$4500a8c0@thomasnotebook>

> This is a huge source of confusion, and I cant remember the rules any
> more.
> 
> For working code, see the "CEnterLeavePython" class defined in the
> pywintypes module (I can send you the source if you dont have them
> handy).

Yes, that would be nice.

> 
> For a long and protracted discussion about this and much more, see the
> thread-sig archives from around 12 months ago - this was discussed in
> so much detail it made my nose bleed :-)

I have browsed over this. However, I did not find a final solution
in it and I was hoping for a more clear answer.

> 
> Mark.

Thanks,

Thomas Heller




From andy at robanal.demon.co.uk  Sun Nov  7 17:06:35 1999
From: andy at robanal.demon.co.uk (Andy Robinson)
Date: Sun, 07 Nov 1999 22:06:35 GMT
Subject: Printing on a Windows workstation
References: <6wOU3.142$dF2.175@198.235.216.4>
Message-ID: <382af7bf.36598555@news.demon.co.uk>

"Robert Parker" <rparker at bserv.com> wrote:

>Hi All
>    I have waded through the man pages and the functions to try and
>determine what to use to print to a printer using Win32.  I don't know if
>there is a COM that could be used or some other method to access the printer
>API in Win32.  I don't feel excited about having to generate lines of escape
>sequences to accomplish formatting and proper presentation.  Does any one
>have any suggestions?
>
>    As a newbie it sure is helpful having this newsgroup.  Am greatful for
>any assistance.
>
>Robert Parker-)
<plug>
In the forthcoming "Python Programming for Win32", I have a chapter on
printing under Windows.  Broadly I cover three suggestions:
- directly drawing to a printer canvas (Pythonwin does exactly what
you want, just like any other Windows graphics code)
- automating Word
- generating PDF
</plug>
Since it is not out yet, You can find some code examples for the Word
stuff at www.robanal.demon.co.uk/PythonWindowsTutorial.pdf

And you should look at PIDDLE (www.strout.net/python/piddle), which
includes the PDF-generating library.  It's quite easy to write scripts
to do any report you want as a PDF document, although the recipient
has to open it and hit 'print'.  

If you want to write directly to a Windows device context, try the
snippet here:

========simpleprintdemo.py==============================
# basic printing demonstration

import win32ui
import win32con

myprinter = "HP DeskJet 820C Series Printer"   
# get the name from your Printers folder

def print_it():
	dc = win32ui.CreateDC()
	dc.CreatePrinterDC()   # ties it to your default printer
	dc.StartDoc('My Python Document')
	#dc.SetMapMode(win32con.MM_LOENGLISH)
	dc.StartPage()
	# text - near the top left corner somewhere	
	dc.TextOut(110,790, 'Hello, World')  # 1 inch in, 8 up
	# try to draw a box around it - not device-specific
	dc.MoveTo(100,800)
	dc.LineTo(400,800)
	dc.LineTo(400,700)
	dc.LineTo(100,700)
	dc.LineTo(100,800)
	dc.EndPage()
	dc.EndDoc()
	print 'sent to printer'
	del dc

=========================================================
The thing I did not cover, because it works but is too ugly for
real-world reports, is that you can print a text file (or Postscript
file if that is what you have) with a DOS command like
		 copy myfile.txt lpt1:
and of course can automate this with a Python call to os.system().


Hope this helps.

Andy





From fredrik at pythonware.com  Fri Nov 19 05:18:37 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 19 Nov 1999 11:18:37 +0100
Subject: Beginning programming with Python
References: <Pine.LNX.4.10.9911191043080.2313-100000@essle.valdyas.org>
Message-ID: <012401bf3277$72897010$f29b12c2@secret.pythonware.com>

Boudewijn Rempt <boud at rempt.xs4all.nl> wrote:
> You needn't guess about that, I guess ;-). I didn't know there
> was a design team for Python, I though it was Guido's invention.

it is -- the CWI researchers (including GvR) worked
on Python's predecessor ABC.

> But I still wonder about the double == - if I can't assign in
> an if, why insist on ==?

probably because binding and comparision
are two different things; consider:

    a = b = c

vs.

    a = b == c

(if we'd start over from scratch, I'd vote for
using ":=" for binding, and "=" for comparision,
but I suppose even more people would have
complained about that one...)

fwiw, ABC used "PUT value IN variable"...

</F>




From bernhard at alpha1.csd.uwm.edu  Tue Nov 16 15:19:29 1999
From: bernhard at alpha1.csd.uwm.edu (Bernhard Reiter)
Date: 16 Nov 1999 20:19:29 GMT
Subject: Beginning programming with Python
References: <sng0ychgwt.fsf@motorola.com> <382B3A53.5816B2@callware.com> <382B3D79.74CD3E0F@email.sps.mot.com> <80p9ds$ppo$1@newshost.accu.uu.nl> <asteroidal-80psfo/INN-2.2.1/bat@broadway.news.is-europe.net> <38312D55.A36EB56D@iqsoft.hu> <antigorite-80rol2/INN-2.2.1/attentive@broadway.news.is-europe.net> <383195C1.826B3342@iqsoft.hu>
Message-ID: <slrn833f2h.6jm.bernhard@alpha1.csd.uwm.edu>

On Tue, 16 Nov 1999 18:34:57 +0100, Nemeth Miklos <nemeth at iqsoft.hu> wrote:
>Janos Blazi wrote:

>The only thing I wanted, is to point out a potential shortcoming (if at all)
>of Python -- I am not sure if the lack of compile-time type safety is really a
>badly needed feature.

It is a bit difficult to make a good case here.
The dynamic nature of python as advantages and disadvantes.
One of the latter being the possibility to have simple bugs in code
which was never executed and not strongly type-tested.

But, this dynamic nature allows for easier an more powerful programming.
Someone could pose the hypothesis that the
advantages in faster programming and easier reading will help the
stability of the programs more overal as introducing strong typing.

Because programmer will see mistakes earlier. Review is easier.
You'll find more capable programmers.


>I am neither a physicist nor a mathematician, I am a simple business software
>engineer. I like Python but I'd like to be convinced that Python is not only
>for scientific programming, but also for business programming. 

It is in my opinion.

>I'd like to know the general opinion of the "official" Python community on
>this (I think) important topic. 
How do you have an official community part of a free software project? :)

>As far as I understood, you are absolutely sure that Python is a perfect
>language, I have to trust in it without any hesitation?

Well no, of course.
Your statements were at least partially not perceived as questions.

I tried to make a point towards dynamic languages.
And additional advantage of python is, that the community has reached
the critical mass, universal, available(Free Software), that it is quite 
stable and realtively cross platform. 

Java would have the community but lacks
stability (in the definition) and has a bunch of cross platform issues.

Of course there are a bunch of very interesting languages out there, but
a plot of them just do have a smaller community or are not as simple as 
python (guile and pike come to mind.)

	Bernhard
-- 
Research Assistant, Geog Dept UM-Milwaukee, USA.  (www.uwm.edu/~bernhard)
Free Software Projects and Consulting 		         (intevation.net)  
Association for a Free Informational Infrastructure            (ffii.org)



From cfelling at iae.nl  Sun Nov 28 17:10:15 1999
From: cfelling at iae.nl (Carel Fellinger)
Date: 28 Nov 1999 23:10:15 +0100
Subject: automatic "lotjes trekken"
References: <000201bf37f4$cce6f820$202d153f@tim>
Message-ID: <81s987$h2h$1@vvs.superst.iae.nl>

Tim Peters <tim_one at email.msn.com> wrote:

> Here's an example showing why you should *not* take the other approaches
> that have been suggested:
...example and explanation snipped...

aiaiaia, mio stupido he. glad my kids aren't lurking on this newsgroup yet;
otherwise they would have argued for a redraw. Anyhow, now I'm forced to
redo the programming. Thanks for sharing all the insights!


> Here are three approaches that work:

> 1) A variant of the one you're already using:  generate a *permutation* at
> random, in one gulp.  Then check whether it meets the restrictions.  If it
> doesn't, try again.  Then each solution clearly has an equal chance of
> getting picked.

> Deep flaw #1:  As explained here a few months ago, Python's random-number
> generator isn't capable of generating more than about 1E12 distinct
> permutations, so even for a list of size 20 it's incapable of generating
> "almost all" of the possible permutations.

I was aware of this one, thanks to the list being a fine tutor.

> Deep flaw #2:  As with your current approach, it will run "forever" if there

the solution as i posted bailed out after 1000 tries, not exactly forever:)

> is no solution.  A variant of the birthday paradox kicks in here too:  even
> if all permutations could be generated at random, "at random" means you get
> duplicates.  The expected number of permutations you have to generate at
> random before you see each of the N! possibilites at least once is about N!
> * ln(N!) (ln == log to the base e).  Cut off the search earlier than that,
> and chances are better than half that you never saw at least one permutation
> (which may have been a solution).

But I have to rethink this one. I'm not trying to generate all solutions,
just one. So in what way would this make the algoritme unfair? Each time
I call this program is an event on its own, unrelated to the previous. So
the 'missing' solution might well be a different one in each run.


> 2) The approach below:  generate all possible solutions, then simply apply
> random.choice to the list of results.  For your size of problem, it will run
> quickly, and in any case runs faster the more restrictions there are.

A variation of this would be to number the solutions, draw a number and
generate that solution as I proposed in a previous post. I will use your
info and some rainy days to figure out whether it is feasable at all.
Would circumvent the next deepflaw too.

> Deep flaw:  For larger problems, it's utterly impractical (too many
> solutions).

> 3) Go back to the theory:  at each step, analyze the board according to the
> method sketched yesterday, and put a rook on that step's distinguised square
> with probability given by the number of solutions that have a rook on that
> square divided by the total number of solutions (i.e., with or without a
> rook on that square).  In the example above, this means that if your first
> choice is wrt the Carel->Martijn square, you would first compute that there
> are 2 solutions with a rook there and 1 without, and so you should put the
> rook there with probability 2/3.  That's the way to do this properly without
> backtracking -- and if you think it's worth the effort, your winter is far
> too long <wink -- but it is a difficult programming problem!>.

still opting for my own variation of 2, but if that fails, I'll try 3.


-- 
groetjes, carel



From praschl at bigfoot.com  Sat Nov  6 21:41:43 1999
From: praschl at bigfoot.com (Michael Praschl)
Date: Sun, 07 Nov 1999 02:41:43 GMT
Subject: open( "file","mode") difference between script and commandline?
Message-ID: <941942257@bigfoot.com>

hi all 

im using python 1.5.2 for BeOS and im having a problem with file
opening. when i use 

>>> f = open( "filename","w" ) 

in the command line interface, this works perfectly ok. but when i use
the same function from a script, i get the following error: 

Traceback (innermost last): 
  File "Signmail.py", line 19, in ? 
    f = open( "/tmp/workfile", "w" ) 
TypeError: illegal argument type for built-in operation 

can anybody tell me why this is? and what i can do about it? 

the second problem i found while i was trying to get around this. i have
a multiline string (linefeeds) in the variable reply, and want to do the
following: 

system( "echo >/tmp/pgpBeMail `reply`" )  #reply is within this "\"
apostrphys (in case the newsreaders cant show them correctly) 

in this case the file is empty after the call. i tried also turning to
this "/" apostrophys and to normal ones, but then the file contains
"reply" as expected, but not the contents of "reply" 

if anybody knows what im doing wrong, please give me a hint! 

-- mfG, Michael Praschl




From fredrik at pythonware.com  Sat Nov 13 06:51:46 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 13 Nov 1999 12:51:46 +0100
Subject: Tkinter Label Widget and the TextVariable.
References: <tIWW3.5785$dz1.510197@pm02news>
Message-ID: <01cf01bf2dcd$76d85d60$f29b12c2@secret.pythonware.com>

Dave Rose <dave.rose at wcom.com> wrote:
> I have assign a textvariable in the instantiation of a series of Labels.  If
> I run this class by itself, the textvariables are displayed.  If I integrate
> the class into a parent application, the text variables do not display even
> though the variables can be printed and do hold the correct variables any
> ideas?

probably a scoping issue, but it's hard to point
out exactly what's wrong unless you post some
sample code.

</F>

<!-- (the eff-bot guide to) the standard python library:
http://www.pythonware.com/people/fredrik/librarybook.htm
(news: just posted an errata for the 'first printing') -->




From parkw at better.net  Thu Nov 11 06:47:36 1999
From: parkw at better.net (William Park)
Date: Thu, 11 Nov 1999 06:47:36 -0500
Subject: How do I compile NumPy on Linux?
In-Reply-To: <382A8094.3354DE70@mdmlab.mi.infm.it>
References: <382A8094.3354DE70@mdmlab.mi.infm.it>
Message-ID: <19991111064736.A528@better.net>

On Thu, Nov 11, 1999 at 09:38:45AM +0100, Sandro Ferrari wrote:
> I downloaded the LLNLDistribution.tgz file and followed the installation
> instruction:
> cd Numerical
> ./makethis.py
> but I get the message:
> make: Makefile.pre.in: No such file or directory
> make: *** No rule to make target `Makefile.pre.in'.  Stop.
> make: *** No targets.  Stop.
> In the documentation isn't mentioned anything about it.
> Any help?
> thanks
> Sandro

This was long time ago.  But, from my memory, I did
    python makethis.py
    python installthis.py
So, check the *.py files whether this should be.



From kanzow at gmd.de  Mon Nov 29 13:33:37 1999
From: kanzow at gmd.de (Peter K.)
Date: Mon, 29 Nov 1999 19:33:37 +0100
Subject: Zope 2: External objects?
Message-ID: <3842C700.A2685499@gmd.de>

Hi, a short question: How do I link external objects,
defined in a Python module, into Zope 2?

Thanks, Peter




From a.eyre at optichrome.com  Tue Nov  2 09:21:23 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Tue, 2 Nov 1999 14:21:23 -0000
Subject: General q. `bout subclassing
In-Reply-To: <381EE780.EBF18F59@bibsyst.no>
Message-ID: <001201bf253d$8ab195b0$3acbd9c2@peridot.optichrome.com>

> How do I subclass fundamental stuff like lists etc,

You can't. But you can subclass a UserList (I don't like this either.)

>>> from UserList import UserList
>>> class my_list(UserList):
...     pass
...
>>> l = my_list()
>>> l.append(1)
>>> l.append(2)
>>> l.append(3)
>>> l
[1, 2, 3]
>>> ^D

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From aa8vb at yahoo.com  Mon Nov  1 11:47:16 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Mon, 1 Nov 1999 11:47:16 -0500
Subject: Doing network traffic in the 'background' from Tkinter app??
In-Reply-To: <ksu2n95191.fsf@telelogic.se>
References: <kspuxy62ub.fsf@telelogic.se> <19991029085829.A106971705@vislab.epa.gov> <ksu2n95191.fsf@telelogic.se>
Message-ID: <19991101114715.B134036529@vislab.epa.gov>

Andrew Markebo:
 |/ Randall Hopper <aa8vb at yahoo.com> wrote:
 || [...] 
 || I can't tell exactly what you're doing from your description, but check
 || out Tk's fileevent callback:
 |
 |Bingo, Thanx!
 |
 |Basically what I want is to give the user the possibilty to mess
 |around with my application while it fetches data from THE internet(TM)
 |
 |Darn, it seems to be Unix only... fungus.. I would like my application
 |to run on windows also..

No first-hand experience mind you, but Tcl newsgroup postings seem to
indicate it works on at least some species of MSWindows:


http://www.deja.com/=dnc/[S0=902331011915dac,ST_rn=ps]/getdoc.xp?AN=519666679

 |I know that there was a bug in Tcl/Tk's fileevent command for Windows in
 |all releases before 8.1.

http://www.deja.com/=dnc/[S0=902331011915dac,ST_rn=ps]/getdoc.xp?AN=503202540

 |Fileevent behavior on Windows is best with the Tcl 8.2 beta, I belive,
 |but workable with wish81.exe.


Randall



From a.eyre at optichrome.com  Tue Nov 23 11:56:34 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Tue, 23 Nov 1999 16:56:34 -0000
Subject: Python and COM: a bug?
In-Reply-To: <816010E2456BD111A48700805FBBE2EEFDF819@ex-quebec-u1.baan.com>
Message-ID: <001e01bf35d3$b2c8f830$3acbd9c2@peridot.optichrome.com>

>> We have had similar problems here with PythonCOM, and have had to stop
>> using it, finding its behaviour too random.

> OK, at least I'm not alone. I wish the problem was easy to reproduce :(

If I had the time and the COM knowledge, I'd debug the source to see what's
going wrong, but I don't.

If this is vital to your project, it might be worth running the source
through a debugger.

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From a.eyre at optichrome.com  Tue Nov 30 04:31:38 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Tue, 30 Nov 1999 09:31:38 -0000
Subject: Python bugs
In-Reply-To: <14402.58513.737532.298175@dolphin.mojam.com>
Message-ID: <001301bf3b15$b391c4d0$3acbd9c2@peridot.optichrome.com>

>     >>> s = shelve.open("foo")
>     >>> l = []
>     >>> s["bar"] = l
>     >>> s["bar"]
>     []
>     >>> l.append("baz")
>     >>> s["bar"]
>     []
> 
> I think you're asking for magic that really can't happen (at least not
> easily).

I understand the problems involved. The odd thing is, because of the way a
shelve is presented to the user (as a dictionary), I just assumed it would
work. Perhaps making it look like a dictionary is a Bad Thing (tm). Okay,
so it is possible to fix, but it would no doubt be highly ineffiecient.
Unless someone else can think of an elegant solution for this one.

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From m.faassen at vet.uu.nl  Mon Nov  1 06:34:04 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 1 Nov 1999 11:34:04 GMT
Subject: Comparison between Python and "Ruby"
References: <3819BCDA.AC4FD4CA@strs.co.uk> <87k8o2x67f.fsf@ev.netlab.co.jp>
Message-ID: <7vjtrc$hmm$1@newshost.accu.uu.nl>

Yukihiro Matsumoto <matz at netlab.co.jp> wrote:
> Ian Clarke <I.Clarke at strs.co.uk> writes:

> |Problems with Python are:

> They are intended to be differences, not problems, OK?

Ok. :) Ruby sounds cool enough. It's received attention from several
Pythoneers in the past, I think, looking at its implementation of classes
for everything, and at its garbage collection system.

>From what I hear in this thread its C interfacing would also be interesting
to compare with Python's, especially since Ruby uses garbage collection,
which is something Python's looking at using in the future, in some
way.

> Python is good enough to satisfy many, but some (including me, the
> designer of the Ruby language) perfer Ruby.  TMTOWTDI.

Agreed. This isn't a flamewar anyway. Ian Clarke chose a far more civilized
title to his post than he did a while back in comp.lang.perl. ;)

What-does-TMTOWTDI-stand-for?-ly yours,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?



From kfitz at shout.net  Mon Nov 29 18:15:41 1999
From: kfitz at shout.net (Kelly Fitz)
Date: Mon, 29 Nov 1999 15:15:41 -0800
Subject: embedded interpreter in shared library
Message-ID: <3843091D.427526E7@shout.net>

I am trying to create a shared object (for IRIX, today, eventually for
Linux and windoze) containing an embedded Python interpreter. My
application loads certain modules (not Python modules, its own sort of
modules) dynamically, and I'd like to provide a dynamically loadable
interface to a Python interpreter. Moreover, I'd like to be able to use
this shared object on machines without a Python installation, so I want
it to be self-sufficient, so to speak. 

My idea was to write a python script that loads the python modules I
want to include, and then freeze this script to generate the necessary c
files for those modules, and then compile and link those into my shared
library. Naturally, it doesn't seem to be working.

I have already created a shared object with an embedded interpreter, but
it requires a Python installation. I am also using SWIG to wrap some of
my own C++ classes into Python modules.

Does anyone know of a more straightforward way to embed an interpreter
in a shared object? Any advice will be welcome. Documentation on
embedding the Python interpreter seems scant at best.

-kel

-- 
---------------------------------------------------------------
 Kelly Fitz                 http://datura.ece.uiuc.edu/Kelly
               member - CERL Sound Group
  research programmer - NCSA Audio Development Group, UIUC
  visiting researcher - WSU Imaging Research Lab



From fredrik at pythonware.com  Wed Nov 10 12:49:17 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 10 Nov 1999 18:49:17 +0100
Subject: statics everywhere.  Why?
References: <80cam4$phe$1@rtpnews.raleigh.ibm.com>
Message-ID: <011701bf2ba4$1a899b80$f29b12c2@secret.pythonware.com>

Arinte <shouldbe at message.com> wrote:
> Embedding python in a c application, why do I need
> to make the functions that python call and variables
> for callbacks static?

beats me.  what happens if you don't
make them static?

</F>




From fredrik at pythonware.com  Thu Nov 18 05:31:32 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 18 Nov 1999 11:31:32 +0100
Subject: Minimal Python installation?
References: <MPG.129cf032cc39a31c989686@news.ericsson.se>
Message-ID: <012a01bf31b0$4e0181f0$f29b12c2@secret.pythonware.com>

Svante Kleist <Svante.Kleist at edt.ericsson.se> wrote:
> We intend to bundle Python with a product, but would like to avoid
> wasting more space than necessary. Is this a problem, license wise?
> I mean, is one supposed to pass the _whole_ 
> Python distribution only?

the license is pretty clear on this: "Permission to use, copy,
modify, and distribute this software and its documentation
for any purpose and without fee is hereby granted" etc.

according to http://www.python.org/doc/Copyright.html,
some lawyers seem to think that "a, b, c, and d" means "a
or b or c or d, but not any other combination", but the
clarification on that page should be enough for your own
company lawyers (if they object, get a second opinion ;-)

</F>




From zych at pool.informatik.rwth-aachen.de  Mon Nov  1 07:10:44 1999
From: zych at pool.informatik.rwth-aachen.de (Gregor Zych)
Date: 01 Nov 1999 13:10:44 +0100
Subject: pipes and commands
References: <87d7ty4cyg.fsf@python.kullen.rwth-aachen.de> <c75aep1n5qq.fsf@pc142.cosc.canterbury.ac.nz> <87904konsk.fsf@python.kullen.rwth-aachen.de> <m3aep0pwfc.fsf@greebo.nodomain.de> <8766znql9i.fsf@python.kullen.rwth-aachen.de> <m33durktvv.fsf@greebo.nodomain.de>
Message-ID: <87iu3m76pn.fsf@python.kullen.rwth-aachen.de>

The problem turned out to be pnmcut(!!!). It seems not to close correctly. 
Thank you very much, guys
	Gregor



From bmcd at es.co.nz  Tue Nov 23 19:14:13 1999
From: bmcd at es.co.nz (Ben Caradoc-Davies)
Date: 24 Nov 1999 00:14:13 GMT
Subject: Method objects question
References: <slrn83lvr7.e7.bmcd@ripley.sulaco.net> <slrn83ku3j.2je.bmcd@ripley.sulaco.net> <slrn83krsu.2i0.bmcd@ripley.sulaco.net> <383a7309@news.uni-bielefeld.de> <81eha1$rf6$1@nntp6.u.washington.edu> <81f4a1$o62$1@nntp6.u.washington.edu>
Message-ID: <slrn83mbr6.ma.bmcd@ripley.sulaco.net>

On 23 Nov 1999 22:26:09 GMT, Donn Cave <donn at u.washington.edu> wrote:
>Well, it's possible, though perhaps not ideal if this were a cornerstone
>of your design.  You just have to add the function first to the class
>namespace, by whatever name is convenient;  then, you can add another
>reference to it in the instance by the name you want, hiding the original
>class attribute.  As long as the assignment that creates the second
>reference locates the function through the instance, you get a bound
>method.
>
>e.g., given class X and instance x,
>  x.bindme = x.bindme_new_7       # bound to instance x
>  x.bindme = X.bindme_new_7       # unbound
>
>So you're forced to use that class name space, but it's accessible
>and if you're worried about collisions, some system can presumably
>be invented to avoid them if you're desperate enough.

My suggestions are already enough of a hack, so I won't worry about collisions!
What you describe is quite a nice trick - binding an instance variable not to a
function object but to a method object. Now the method object is created when
the variable is bound, not when it is accessed. This certainly works around the
problem I described.

The only trap is that if the new function (bound in the class namespace) is
replaced, the binding in the instance must also be replaced, because the method
object still references the first new function.

Hey! Doesn't this solve the namespace collision problem? Now we can delete the
new function from the class namespace - it will live on as a reference in the
method object to which the instance variable is bound.

[Runs off and checks whether this works ...]
 
Yes! Now my problem is solved entirely with no adverse effects or strange
variables lying around.

Thanks, Donn!

Here's the example (compare with the start of this thread);

#################################################

class Demo:
  def __init__(self):
    self.i = 0

a = Demo()
b = Demo()

def incr_i(self): 
  self.i = self.i + 1

def decr_i(self): 
  self.i = self.i - 1

Demo.temp_foo = incr_i     # Here is Donn's technique.
a.foo = a.temp_foo         
del Demo.temp_foo          # Now the method has been safely constructed, we can
                           # restore the class Demo to it's original state.
Demo.temp_foo = decr_i     # And repeat ...
b.foo = b.temp_foo
del Demo.temp_foo

a.foo()                    # Just as desired!
b.foo()

print a.i                  # Correct results.
print b.i

#################################################

>I sure don't know how anything gets decided.  But I'm pretty sure I've
>used this the other way - that is, where the newly added function really
>is just a function, not a method.  I don't think that's terribly exotic,
>and it's possible due to the way this works.

And if instance functions were converted to methods, you would have to write
a wrapper for every ordinary function you wanted to store in an instance.

Thanks again.

-- 
Ben Caradoc-Davies <bmcd at es.co.nz>



From sessile at in-gen.net  Mon Nov  8 21:12:17 1999
From: sessile at in-gen.net (sessile at in-gen.net)
Date: Mon, 08 Nov 1999 21:12:17 -0500
Subject: Can't use len() on returned PyList object?
Message-ID: <3.0.6.32.19991108211217.007cb210@mail.in-gen.net>

Why can't I use 'len()' to find the length of a list
returned from my new extension module?  The error is
shown as:

    TypeError: call of non-function (type int)

The list was created by 'PyList_New()' and populated
using 'PyList_SetItem()'.  I get the same error message
regardless of the data type held in the list.  I can
access each element of the list just fine, but I don't
always know how many there are.  Is there a reasonable
way for me to find the length of these lists in Python
other than returning the length from the extension as a
separate object?  

Python-1.5.2
SunOS-5
gcc-2.7.1

Thanks.




From ionel at psy.uva.nl  Thu Nov  4 04:32:18 1999
From: ionel at psy.uva.nl (Ionel Simionescu)
Date: Thu, 4 Nov 1999 10:32:18 +0100
Subject: how to use pythondoc on Win32?
Message-ID: <7vrjqb$aju@mail.psy.uva.nl>

Hi,

I've tried for about an hour to unwind the web of smart imports used by
pythondoc without any success.

If anyone has a working installation, please let me have a copy of the
critical files
or just tell me what needs to be patched.


Thank you,
ionel


PS
The pythondoc dir is on my path,
I did rename the extensionless 'pythondoc' to 'pydoc.py',
but
I get an import error in pythondoc/xref/xref.py : function main(...)
It cannot import internalXref, then it cannot import message.









From paul at digicool.com  Wed Nov 10 06:49:22 1999
From: paul at digicool.com (Paul Everitt)
Date: Wed, 10 Nov 1999 06:49:22 -0500
Subject: FYI: Python company worth $100M
Message-ID: <38295BC2.13E2E31F@digicool.com>

Next time someone claims that Python is marginal...next time that
someone says nobody uses Python...

eGroups is merging with Onelist (a competitor), and each is valued
around a million bucks:

  http://www.thestandard.com/articles/display/0,1449,7543,00.html

For those that don't know, eGroups is a community website.  At the
O'Reilly Conference, Scott Hassan (looonnngg time Python star) described
the technology they use.  It's a massive farm of machines with
applications all written in Python.  In fact, they bailed on sendmail,
then bailed on qmail, and rewrote their mail delivery system in Python. 
Scott also said that the big win came from Sam Rushing's asynch
architecture, part of which is now in the Python standard library.

Nice guys _don't_ finish last, it seems.  Congrats to Scott and Sam, and
yippee for us, because this is good news!

--Paul



From robin.boerdijk at nl.origin-it.com  Thu Nov 25 15:47:19 1999
From: robin.boerdijk at nl.origin-it.com (Robin Boerdijk)
Date: Thu, 25 Nov 1999 21:47:19 +0100
Subject: creating an AIX shared object for Python
References: <81haen$5qg$1@nnrp1.deja.com>
Message-ID: <0E16861EE7BCD111BE9400805FE6841F0CC35C22@c1s5x001.cor.srvfarm.origin-it.com>

In Python-1.5.2/Misc there is a file called AIX-NOTES which has a section
'SHARED LIBRARY SUPPORT' and that explains how to do it.

Robin

<drjatalb at my-deja.com> wrote in message news:81haen$5qg$1 at nnrp1.deja.com...
> My question is in trying to create a shared object that can be
> dynamically loaded into Python on AIX.  I have a small extension module
> of C code which makes a 3rd party C library available to Python.  I
> have been successful in creating a shared object on IRIX, Solaris, HP-
> UX, and Windows NT for this extension module but so far have not been
> successful in doing it under AIX.  Here are the two simple commands I
> am trying to use and the error message which results during the build
> of the shared library (I am using Python 1.5.2):
>
> cc -c -qfuncsect -I/users/toollib/include/python1.5 -
> I/net/rush/disk1/pdi/toolkits/crusher/crusher312/include crushermodule.c
>
> ld -G crushermodule.o -o crushermodule.so -
> L/net/rush/disk1/pdi/toolkits/crusher/crusher312/lib/AIX_4.2.1 -lcrusher
> ld: 0711-327 WARNING: Entry point not found: __start
> ld: 0711-244 ERROR: No csects or exported symbols have been saved.
>
> Any suggestions would be appreciated.
>
> Dr. Jeff Talbert
> jat at iti-oh.com
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.





From a.eyre at optichrome.com  Wed Nov 24 04:19:53 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Wed, 24 Nov 1999 09:19:53 -0000
Subject: Block and continue question
In-Reply-To: <81ff6o$rrg$1@nnrp1.deja.com>
Message-ID: <002801bf365d$10ed7880$3acbd9c2@peridot.optichrome.com>

> What is the most elegant way to do this within
> Python?

Use an extension module.

Read the chapters about extension modules on the
Python website. But basically, if the C function
blocks, then so will the Python function.

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From aahz at netcom.com  Thu Nov 11 10:46:23 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 11 Nov 1999 15:46:23 GMT
Subject: What are Aspects? (Re: How can I make my assertions smarter?)
References: <Pine.WNT.4.05.9911032042440.109-100000@david.ski.org> <7vvq63$ehm$1@nntp4.atl.mindspring.net> <3826C32A.FB47EF6@compaq.com> <slrn82h5o6.26j.wtanksle@hawking.armored.net>
Message-ID: <80eocf$8rc$1@nntp5.atl.mindspring.net>

In article <slrn82h5o6.26j.wtanksle at hawking.armored.net>,
William Tanksley <wtanksle at hawking.armored.net> wrote:
>>Aahz Maruch wrote:
>>>
>>> Sounds like templates to me, or Python's "mix-in" classes.  Am I missing
>>> something?
>
>The _primary_ idea, though, is that aspects can contain functionality
>which classes simply can't.  In order to implement reference counting with
>classes, for example, you have to modify tons of source code by hand, all
>in seperate classes.  If you use aspects, though, you can start with code
>which has _no_ memory management, and write a single aspect which
>describes reference counting, and have most of your memory management
>problems solved.  (A second aspect might help with cycles, and yet another
>aspect to handle some memory management staticly, since in some cases it's
>easy.)

I feel like I'm still missing something.  What if you only want half of
your classes to get ref-counting?  Or are you saying that aspects are by
definition applied across *all* code, and that they're more like hooks
into the Python interpreter?
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From jeske at egroups.net  Sun Nov 21 21:00:43 1999
From: jeske at egroups.net (David Jeske)
Date: Sun, 21 Nov 1999 18:00:43 -0800
Subject: Python Type-Inference based LINT.. (pylint.py)
Message-ID: <19991121180043.B3184@teapot.egroups.net>

Hello,

I'm interested to know who out there is interested in working on a
type-inference based static type checker (i.e. LINT) for Python. I've
read various snippits about static type extensions, and the intent to
put type-inference into Viperi. However, I've not seen much more talk
of either of these.

I've been thinking that even with syntax extensions like the
"Add-Later Static Types" proposal, type-inference would be required to
get any valuable static type-checking out of the majority of (untyped)
Python code. Furthermore, a Type-Inference based static-lint should
allow some static type safety for those who want it, without impacting
any the existing Python environment for those who don't want it.

Scott Hassan and I have started a simple type-inference graph
generator based on the parser module's Abstract Syntax Trees. It is
able to parse basic test files and produce some simple type-inference
graphs. We could use some help from someone more familiar with
both ASTs and the Python Interpreter Internals.

You can download what we have so far by by going to:

http://www.chat.net/~jeske/Projects/PyLint/download/

The source file includes our theory of operation, implementation
documentation, and of course, the code. :)

-- 
David Jeske (N9LCA) + http://www.chat.net/~jeske/ + jeske at egroups.net
                    eGroups Core Server Engineering



From aa8vb at yahoo.com  Wed Nov 24 09:34:41 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Wed, 24 Nov 1999 09:34:41 -0500
Subject: What does Tkinter do when Idle ?
In-Reply-To: <3832DCC6.60E05050@lri.fr>
References: <3832DCC6.60E05050@lri.fr>
Message-ID: <19991124093441.A1426604@vislab.epa.gov>

Stephane Conversy:
 |It seems that from time to time (say every 10 s.) , Tkinter wakes up
 |and do something, and then comes back to idle.
 |
 |is it right ?

What OS?  MSWin?  or a UNIX varient?

What clues do you have that something is waking up?  What indicates that
this is your Python process?  Do you have any clues this is Tkinter, or Tcl
specifically, or a callback you've installed?  What Tkinter callbacks has
this process installed?  Any timers?  Is it sufficiently short (< 50 lines)
to post?

More input Stephane!  Need more input! ;-)

(I can tell you I don't see behavior in general.)

-- 
Randall Hopper
aa8vb at yahoo.com



From ivanlan at callware.com  Fri Nov  5 13:18:46 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Fri, 05 Nov 1999 11:18:46 -0700
Subject: eff-bot's The Standard Python Library
References: <19991105125124.A302546@vislab.epa.gov>
Message-ID: <38231F85.10CF427C@callware.com>

Hi All--

Randall Hopper wrote:
> 

[bobbitt]

>      Install note: No luck running it on NT4; no GUI or console messages
> appeared at all, though I didn't have Acroread installed there (possibly
> what confused it, though I would still have expected an error message). 

I have NT4.0 SP5; the Acrobat reader installed correctly, even told me
that Acrobat 3.0 was installed and that I should uninstall that.  I
chose not to, because I have a licensed copy of the Acrobat Exchange pdf
generator and so on along with 3.0.  

I did have to go in and edit the path to the Acrobat executable in file
associations, though, in order to get pdf files to be read by 4.0.

<my-nt-is-smarter-than-yours>-ly y'rs,
Ivan;0)
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
----------------------------------------------



From flowit at hotmail.com  Sat Nov  6 15:57:06 1999
From: flowit at hotmail.com (Florian W.)
Date: Sat, 6 Nov 1999 21:57:06 +0100
Subject: Doing it with Python, Windows SEND-TO and *.cpl-files
Message-ID: <8024lh$2mnb$1@bnews.gigabell.net>

Hello Python-people

I'm working with Ms Windows.

When I push a script-into the Send-To-Folder,
what must be done, to get the filename which was selected with it ?

Is it possible to create a *.cpl-file ?

What is the easiest way to manage dialogs ? (With dialog-boxes and all
rc-file-stuff)

Mfg Florian






From m.faassen at vet.uu.nl  Thu Nov 25 05:16:56 1999
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 25 Nov 1999 10:16:56 GMT
Subject: automatic "lotjes trekken"
References: <81cqnv$2ip$1@vvs.superst.iae.nl> <81e3s2$j14$2@newshost.accu.uu.nl> <81e41d$j14$3@newshost.accu.uu.nl> <383BC208.4795B8CB@compaq.com>
Message-ID: <81j2ao$nh2$1@newshost.accu.uu.nl>

Greg Ewing <greg.ewing at compaq.com> wrote:
> Martijn Faassen wrote:
>> 
>> So,
>> each time a person is drawn by someone, the program removes that program
>> from everybody else's lists (if that person was in there in the first
>> place).

> That doesn't guarantee that everyone gets drawn, though.
> You might get down to drawing for the last person and find
> that there's nobody left on their list, even though there
> are people remaining on other lists that have never been
> drawn.

Aaah, right. After some more pondering I see what you mean now. Actually
I whipped up a script and I almost posted it until I figured out what
could go wrong. :)

> I can't think of any method that wouldn't have to backtrack
> under some circumstances.

Right, at least apparently it's not as straightforward as I thought. 
Backtracking could at least be made very unlikely if you draw for the
person that has the fewest options left first, each time. But I think
even that strategy could end up in a deadlock situation on occasion.
 
Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?



From gerrit.holl at pobox.com  Tue Nov  2 16:44:36 1999
From: gerrit.holl at pobox.com (Gerrit Holl)
Date: Tue, 2 Nov 1999 22:44:36 +0100
Subject: capturing stderr && stdout
In-Reply-To: <Pine.LNX.4.10.9911021350390.15744-100000@jazz.monsterbymistake.com>
References: <Pine.LNX.4.10.9911021350390.15744-100000@jazz.monsterbymistake.com>
Message-ID: <19991102224436.A12442@optiplex.palga.uucp>

Agent Drek wrote:
> Hullo,

Hyllo,

> p = os.popen("%s" % HSCRIPT, "w")
> p.write("opcd /obj\n")
> p.write("opadd geo\n")
> p.write("opls\n")
> p.write("quit\n")

Hmm, I recommend replacing this with:
p.writeline("opcd /obj")
p.writeline("...")
...etc...

or:
p.write("""opcd /obj
opass geo
opls
quit""")

or:
STRING = """opcd /obj
opass geo
opls
quit"""

p.write(STRING)

regards,
Gerrit.



From krussll at cc.umanitoba.ca  Tue Nov  2 17:46:12 1999
From: krussll at cc.umanitoba.ca (Kevin Russell)
Date: Tue, 02 Nov 1999 16:46:12 -0600
Subject: Audio Mixer for win32
References: <sj8T3.2314$Pf3.129511@news.uswest.net>
Message-ID: <381F69B4.763F895B@cc.umanitoba.ca>


shdr wrote:

> I've been looking through all the python documentation I could find as well
> as hundreds of other resources for a way to change the audio mixer levels in
> python. I simply want to write a python script that would allow me to change
> the mixer volumes. On a similar note I would also like to be able to record
> from the mic or line-in inputs. If anyone has any experience or expertise to
> share with me it would be much appreciated.

I've never tried this, but one possibility would be to work through one of
the TCL extensions for sound -- I'm thinking specifically of Snack, but
there are probably others.  On the plus side, your results would be more
portable than if you manipulated the guts of Microsoft directly.  On the
down side, it involves interfacing with an inferior language. :-)

Python can do some interesting things to sounds once they're in memory
or a file, but for getting them there and sending them out again its abilities
are rahter -- um -- minimal.  It's kind of ironic, given that the very first
name on the list of "invaluable contributors" to SoX is a certain Dutch
guy named Guido...

Audio SIG, anyone?

-- Kevin





From eric at fruitcom.com  Mon Nov  1 01:39:28 1999
From: eric at fruitcom.com (Eric Smith)
Date: 1 Nov 1999 06:39:28 GMT
Subject: multiline prototyping on command line (oops)
References: <slrn81la5n.2fh.eric@plum.fruitcom.com> <87so2ttegm.fsf@slip166-72-251-9.ma.us.prserv.net> <slrn81lf9o.2fh.eric@plum.fruitcom.com> <01111999.2@sanctum.jae>
Message-ID: <slrn81qdcv.lr2.eric@plum.fruitcom.com>

Juergen -
jae at ilk.de (Juergen A. Erhard) posted in comp.lang.python on Mon, 
01 Nov 1999 03:29:37 +0100:
 > 
 > I usually either run a python session in an xterm (bad for multiline), or
 > do python -c '...' multiline.  In zsh  ;-)
 > 
 > Really, try the zsh... I with readline were as good as zsh's ZLE.
 > Kick-Ass (really) multiline editing.  And a history that correctly
 > deals with it.
 > 
 > You can also edit here-documents...
<snip>
 > 
 > Go back in history, and edit the little snippet... no problem.

That is the best answer I have seen.  The only real simple solution.  As
you say if you use the -c, then because of the fact that py uses
indentation, you have to specially insert extra quoting to help the compiler
interpret correctly.

I now have _no_ excuse for proceeding with my long delayed migration to zsh
_and_ to start working with python.

Thanx!
-- 
Eric Smith
eric at fruitcom.com
www.fruitcom.com

I truly believe computer programmers want to be creative, and they may have
many different reasons for wanting to write code a particular way.  What you
choose to optimize for is your concern, not mine. I just supply the paint.



From mwh21 at cam.ac.uk  Wed Nov 24 14:22:57 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 24 Nov 1999 19:22:57 +0000
Subject: newbie question
References: <943469319.838266356@news.teleweb.at>
Message-ID: <m3u2mbsnhq.fsf@atrus.jesus.cam.ac.uk>

Markus Krutz <krutz at cg.tuwien.ac.at> writes:

> How the f*** do I create a string containing a *single* "\" ?

"\\"

You might think that doesn't work because of

>>> "\\"
'\\'

but

>>> print "\\"
\

that's-another-one-for-__str__-by-default-methinks-ly y'rs
Michael



From caw at cs.mu.oz.au  Tue Nov 30 18:19:06 1999
From: caw at cs.mu.oz.au (Chris WRIGHT)
Date: Wed, 01 Dec 1999 10:19:06 +1100
Subject: What's the canonical vi setting for Python indentation
Message-ID: <38445B6A.A1283D2E@cs.mu.oz.au>

I've tried to find out the "OK" indentation settings for python with vi,
and a troll through deja-news found some hints (like don't do indent =
4). Yet I seem to also remember some advice to the contrary?? I want
indentation to be 4 characters, and readable by/ acceptable to emacs
python-mode, and to pass tabnanny...

I think that this should go in the faq...(Couldn't find it from
www.python.org search)


cheers
and thanks

chris wright
-- 
Dr. Chris Wright
Deputy Director, Intensive Care Unit
Monash Medical Centre



From david at hotjobs2000.com  Thu Nov 11 15:25:12 1999
From: david at hotjobs2000.com (David Winsen)
Date: Thu, 11 Nov 1999 12:25:12 -0800
Subject: WEB DATABASE PROGRAMMER POSITION AVAILABLE
Message-ID: <MuFW3.102$yd1.9175@typhoon-la.pbi.net>

WEB DATABASE PROGRAMMER POSITION AVAILABLE

URGENT MESSAGE!

From:  David Winsen-Senior Consultant - High Technology Executive Search

We have an out-dated copy of your resume in our database or have viewed your
credentials on the Internet.  HTES is an established national Executive
Search and Consulting Firm who has been serving the High Tech Industries for
over 25 years.

We have been confidentially retained by a Los Angeles; Ca. based Adult
Internet Fulfillment/Billing Company. Salary is $50K-$100K DOE.
We are confidentially pre-screening top candidates for the following
position: Web Database Programmer
Description

Candidates will need to be extremely detail oriented and have a solid work
ethic. Will be developing cutting edge software and e-commerce applications.
They are an established and ever growing Internet Fulfillment/Billing
Company. They offer a casual & unique work environment unlike any other,
full benefits, & room for growth and advancement.

Requirements

Candidates will need experience in Perl, Python, PHP, Javascript, UNIX,
Linux or FreeBSD, MySQL a plus. BS Computer Science or equivalent. At least
3 years of Web experience is a plus.

If you are interested, please E-mail me in MS Word 95-98 a recent copy of
your resume and a cover letter with your specific information, including
your recent compensation package to Position-for: Web Database Programmer

My personal E-mail is david at hotjobs2000.com or fax your resume to (310)
855-0840. If you have any questions about the position(s), please call me at
(310) 855-0406 and I will discuss them in detail.

We also have developed an interactive Website that you can view over 6000
national openings www.hotjobs2000.com.  This system is effective, easy to
use and new positions are posted daily.  We encourage you to use it and
nominate yourself for other positions you feel you are qualified for.  We
are looking forward to working with you now and in the future.







From greg.ewing at compaq.com  Thu Nov 25 06:16:50 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Fri, 26 Nov 1999 00:16:50 +1300
Subject: Python complaints
References: <000601bf36be$f1bf5e80$292d153f@tim>
Message-ID: <383D1AA2.35E8F0EC@compaq.com>

Tim Peters wrote:
> 
>     mutex.acquire()
>         do stuff
>     mutex.release()
> 
> is a syntax error in Python.

What's *really* needed here is a way to abstract these kinds
of constructs. In Scheme I would arrange things so that I
could write

   (with-mutex-acquired mutex
      some 
      stuff)

or in Smalltalk

   mutex do: [
      some
      stuff
   ]

In Python you have to either put the stuff in a separate
function (which interrupts the flow of thought for both the
writer and reader) or use a lambda (which severely restricts
the form of the stuff). In either case you have scope problems.

So what I'd like is a construct something like

   with_mutex_acquired(mutex):
      some
      stuff

which would call a function defined like

   def with_mutex_acquired(thunk, mutex):
      mutex.acquire()
      try:
         thunk()
      finally:
         mutex.release()

Greg



From gmcm at hypernet.com  Tue Nov 30 14:16:02 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 30 Nov 1999 14:16:02 -0500
Subject: Python Make
In-Reply-To: <38441770.80C33C93@mdlink.de>
Message-ID: <1268133125-35929318@hypernet.com>

Helge Hess wrote:
> 
> Ian Clarke wrote:
> > 
> > Hi, just a quick idea I had that I wanted to share with the
> > newsgroup. Has anyone considered using Python as a replacement
> > for the "make" utility?  A module could be provided which did
> > several small things like testing modification times of a list
> > of source files against a destination file to see whether an
> > update was required.
> 
> YES! I had the same idea :-)

So did someone else. There's one in the contrib area of the ftp 
site, under /System (the larger of the 2 pymake-* downloads).

- Gordon



From aa8vb at yahoo.com  Tue Nov 30 18:04:35 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Tue, 30 Nov 1999 18:04:35 -0500
Subject: Looking for sshlib.py
In-Reply-To: <19991130173837.A2365@quark.emich.edu>
References: <iwI04.9621$VA6.47863@news-server.bigpond.net.au> <19991130080326.A1376@quark.emich.edu> <tBX04.11350$VA6.54975@news-server.bigpond.net.au> <19991130173837.A2365@quark.emich.edu>
Message-ID: <19991130180435.A1935602@vislab.epa.gov>

Jeff:
 |On Tue, Nov 30, 1999 at 09:47:05PM +0000, John Leach wrote:
 |> I'm looking to write a Python frontend for character-based applications
 |> accessed via the internet.
 |> telnet works well but I'd like something with more security (telnet sends
 |> its password unencrypted I believe).
 |> I'll try calling ssh from within Python and see if that helps.
 |> Thanks for your reply,
 |> John Leach
 |> http://www.bosware.com.au
 |> 
 |
 |you are correct in your statement that telnet sends it's passwords in the
 |clear. ssh has clients designed to replace 'rsh' and 'rcp'.

I second the ssh/scp recommendation, however that telnet clients do not
encrypt passwords is an overgeneralization.  Some telnet clients support
password encryption (e.g. via Kerberos or SRP).  No comment on how
secure/insure their encryption methods are, but it's better than
plaintext. ;-)

-- 
Randall Hopper
aa8vb at yahoo.com



From a.eyre at optichrome.com  Mon Nov  1 04:34:16 1999
From: a.eyre at optichrome.com (Adrian Eyre)
Date: Mon, 1 Nov 1999 09:34:16 -0000
Subject: PythonWin Debugger
In-Reply-To: <004401bf2264$135b7ca0$0501a8c0@bobcat>
Message-ID: <001f01bf244c$43f705c0$3acbd9c2@peridot.optichrome.com>

>> win32ui: The frame does not exist

> Hmmm.  The debugger should run from all embedded apps, except Tkinter
> ones (and therefore except IDLE).

Does that include straight win32 (non-MFC) apps?

> This particular problem would appear to be a bug.  I have another
> debugger bug I need to track down, and hopefully I can also solve
> this.

Is it connected with this one, or is it completely different?

> Unfortunately, Tkinter and any other GUI system dont seem to like
> living together.  Eg, I have managed to use the debugger to debug a
> Python COM object when hosted in VB (which is quite a cool demo :-)
> but have never managed to get Tkinter apps working with it.  I would
> really like to, but the little I know about Tk makes it seem somewhat
> difficult.  A Tk expert willing to help me could solve this.

Well, I'm not too worried about the Tkinter thing. I'm only embedding
IDLE until I can get PythonWin's debugger running in my app. I'll give
it another go without Tkinter loaded.

> OTOH, the best stragegy for this is probably remote debugging - use a
> seperate process to perform the debugging.  This would also solve a
> few of the hairy "blocking" problems I need to deal with.  However,
> this doesnt appear trivial either :-(

I should imagine not. I suppose there is some sort of API under Windows
to allow this sort of thing (like the Visual C++ debugger does) but I
wouldn't have any idea where to start.

Cheers...

--------------------------------------------
Adrian Eyre <mailto:a.eyre at optichrome.com>
Optichrome Computer Solutions Ltd
Maybury Road, Woking, Surrey, GU21 5HX, UK
Tel: +44 1483 740 233  Fax: +44 1483 760 644
http://www.optichrome.com 
--------------------------------------------




From s323140 at student.uq.edu.au  Sun Nov 28 10:26:30 1999
From: s323140 at student.uq.edu.au (Rob Hodges)
Date: 29 Nov 1999 01:26:30 +1000
Subject: XPython (X11)
References: <383E5350.1E07536A@xrce.xerox.com> <slrn83tj53.k7h.bernhard@alpha1.csd.uwm.edu> <81ms2n$126$1@essle.valdyas.org> <slrn840dt7.mov.bernhard@alpha1.csd.uwm.edu>
Message-ID: <m3hfi64oyh.fsf@phaedrus.zen.xom>

bernhard at alpha1.csd.uwm.edu (Bernhard Reiter) writes:

> >Hey! Don't forget PyQt and PyKDE (and PyQt seems to be
> >working under Windows, too, nowadays)...
> Do you have to but the Troll Kit first? >:)
> 
> And I didn't mention pyGNOME, why should I mention to PyKDE?
> To create an application tied to KDE or GNOME is a mistake in my
> eyes anyway. <wink>

I'm struggling a bit to understand why you say that in this context.
I can't speak for pyKDE (although I imagine the situation is the
same), but developing with pyGnome doesn't imply a 'tying' to Gnome;
only that the Gnome libs must be available in addition to the GTK libs
-- just as you must have wxwin in addition to GTK to run wxpython
apps.  One advantage of PyGnome over straight PyGtk is that it offers
the Gnome Canvas widget, which can be useful whether the user is
running Gnome or not.

There is also no harm in taking advantage of the various services
provided by these desktop environments to improve the user's
experience, provided that your app also works nicely without them;
that's what Gnome and KDE are all about, after all.  If you do this
stuff and your app *doesn't* work nicely outside, I agree you've made
a mistake -- but in your implementation; not in your choice of
toolkit.

And yeah, last I heard you had to buy the Troll Kit for PyQt under
Win32.  But I'm sure I wasn't hallucinating when someone on the PyGtk
list mentioned having ported PyGtk to Win32... has anybody tried out
the port?  I strongly prefer PyGtk over WxPython on Linux, so I'm
curious in case I should have to develop anything for Win in the
future (but, I guess, not curious enough to reboot :)

-Rob




From prestonlanders at my-deja.com  Mon Nov 29 13:14:10 1999
From: prestonlanders at my-deja.com (Preston Landers)
Date: Mon, 29 Nov 1999 18:14:10 GMT
Subject: Unbuffered stdout
References: <81u1k2$27qkk$1@oceanite.cybercable.fr>
Message-ID: <81ufpe$se3$1@nnrp1.deja.com>

Do you have something equivelent to the following at the beginning of
your output routine?

print "Content-Type: text/html\n\n"

If you don't put this, you won't see anything with most combinations of
servers and browsers.

good luck,

---Preston

In article <81u1k2$27qkk$1 at oceanite.cybercable.fr>,
  "leon" <leonar_n at epita.fr> wrote:
> Hi there !
>
> I'm developping an IRC like software which must work on the Web. I
can't use
> Java,
> and I'm using Python to program the CGI.
> But, I have a big problem : I don't know how to make the stdout
unbuffered.
> I tried to make some
>     sys.stdout.write("test")
>     sys.stdout.flush()
> but it don't work.
>
> So any idea ??
>
> Leon.
>
>

--
|| Preston Landers <prestonlanders at my-deja.com> ||


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



From aahz at netcom.com  Sat Nov  6 12:39:06 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 6 Nov 1999 17:39:06 GMT
Subject: making the Python case to my dept.
References: <9911051740130D.02530@william.jinsa.org> <3823FC43.CF7FBD71@python.net>
Message-ID: <801p3q$q01$1@nntp2.atl.mindspring.net>

First of all, I want to second Patrick's suggestion that this be FAQ'd.
I have one nit, though:

In article <3823FC43.CF7FBD71 at python.net>,
Thomas A. Bryan <tbryan at python.net> wrote:
>
>Still, few people are knocking down my door for Python programmers 
>while there are plenty of jobs for C++/Java/Delphi programmers.

I believe that there are more jobs for good Python programmers than
people with the skills to fill them.  My company is certainly having
difficulties (jobs in Mountain View! ask me!), and having HP Cupertino
advertising for 10-12 Python positions only increases the difficulty.

I think we've come to the knee in the curve where Python has achieved
critical mass and will get a lot of popular attention over the next
couple of years.  Few of us need worry about job security.
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From mgm at unpkhswm04.bscc.bls.com  Thu Nov 18 15:31:58 1999
From: mgm at unpkhswm04.bscc.bls.com (Mitchell Morris)
Date: 18 Nov 1999 20:31:58 GMT
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <80pe77$osc$1@hub.org> <slrn830he9.n28.mgm@unpkhswm04.bscc.bls.com> <80vp2m$275$1@nntp9.atl.mindspring.net>
Message-ID: <slrn838ohs.fv8.mgm@unpkhswm04.bscc.bls.com>

In article <80vp2m$275$1 at nntp9.atl.mindspring.net>, Aahz Maruch wrote:
>In article <slrn830he9.n28.mgm at unpkhswm04.bscc.bls.com>,
>Mitchell Morris <mitchell at morrisland.com> wrote:
>>
>>As an aside, the other thing about Perl that I like is perldoc. As much as I
>>use lynx, it still sucks for searching the Python library reference. At some
                               ^^^^^^^^^
>>point I'll write ``pythondoc'' to extract the parts I care about, but until
>>then I'll just keep kvetching.
>
>I don't have any problems using Lynx with the Python docs.

Reading is not so bad ... searching blows. As an example, I'm looking at the
page for the "re" module and want to see the argument list for "unpack" but
can't seem to remember if it's in "string" or "struct" or
Guido-only-knows-where. Lynx provides squat for this operation, so I wind up
manually traversing the HTML hierarchy until I find it. This is the itch
that I'll wind up scratching sooner or later.

traversing-the-doc-tree-ly y'rs,
+Mitchell

-- 
Mitchell Morris

But I gave up on saying that out loud because a lot of people say that and I
don't want to sound like a lot of people.
	-- George Carlin



From cgw at fnal.gov  Thu Nov 25 06:08:29 1999
From: cgw at fnal.gov (Charles G Waldman)
Date: Thu, 25 Nov 1999 12:08:29 +0100 (MET)
Subject: Python complaints
In-Reply-To: <86puwz4jhp.fsf@g.local>
References: <1268673769-3409210@hypernet.com>
	<86puwz4jhp.fsf@g.local>
Message-ID: <14397.6317.426060.330244@gargle.gargle.HOWL>

Gareth McCaughan writes:
 > Gordon McMillan wrote:
 > 
 > [I said:]
 > >>   - if/then/else works on *statements* (strictly, suites of
 > >>     statements) rather than on *expressions*, so I can't
 > >>     say
 > >>       print "You scored %s point%s" % (score, if score==1 then ""
 > >>       else "s")
 > >>     or anything like that. (Or, as some people prefer to put it,
 > >>     Python has no ? : operator.)
 > > 
 > >>>> def p(score):
 > > ...  print "You scored %d point%s" % (score, score==1 and ' ' 
 > > or 's')
 > > ...
 > >>>> p(1)
 > > You scored 1 point
 > >>>> p(3)
 > > You scored 3 points
 > >>>> 
 > 
 > Yes. Unfortunately, as we all know, this doesn't work when
 > it's (say) 1 and 0 you're choosing between instead of "s" and ""...
 > 

You can often (ab)use dictionaries to get a ?:-like effect, as in this 
construction:

def p(score):
    print "You scored %d point%s" % (score, {1:''}.get(score,'s'))




From krutz at cg.tuwien.ac.at  Tue Nov 23 16:06:02 1999
From: krutz at cg.tuwien.ac.at (Markus Krutz)
Date: Tue, 23 Nov 1999 21:06:02 GMT
Subject: urllib.urlopen-bug?
Message-ID: <943390670.1482784211@news.teleweb.at>

Hi,
I've got a problem using the urlopen-function from the urllib.
According to the library reference, an IOError-exception should be raised on
two conditions:
1) The connection can't be made.
2) The server returns an error.
On 1) an exceptions is raised, but if I do a
urllib.urlopen("http://www.python.org/nonexistingdocument.html")
no exception is raised, although I think this is a typical 404-Error.
In my opinion this should result in raising an exception...
Any hints on other ways how to check for the existence of a net resource?
(I'm using python v1.5.2)

greetings
Markus




From ionel at psy.uva.nl  Tue Nov 30 13:57:00 1999
From: ionel at psy.uva.nl (Ionel Simionescu)
Date: Tue, 30 Nov 1999 19:57:00 +0100
Subject: Does this code create a circular reference ?
Message-ID: <8216ih$2ki@mail.psy.uva.nl>

Hi,

I need advice on whether the situation described below introduces circular
references.

---

Basically, it is about a class that, within some method,
needs to create a temporary "child" object,
that is an object holding a reference to its "creator".

This is intended as a simple way to share access to certain variables known
by the parent.

I guess a danger in this approach, because the child,
even temporary, will be somehow stored in the locals dictionary of the
method.

I am not sure how it works and if it will cause trouble.

---

class child:
  def __init__(self, parent):
    self.parent = parent

  def do_something(self):
    pass


class headache:
  def method(self):
    a_child = child(self)
    a_child.do_something()

---

Thank you,
ionel






From wware-nospam at world.std.com  Tue Nov 23 22:56:39 1999
From: wware-nospam at world.std.com (Will Ware)
Date: Wed, 24 Nov 1999 03:56:39 GMT
Subject: PalmPython?
Message-ID: <FLooAG.MqL@world.std.com>

Is PalmPython still around? I think Bob Tillotson (sp?) had been
working on it. I picked up a Palm Pilot today, and have gotten
pilot-link working on my Linux box. PalmPython would be a cool
next thing to get working, if it still exists. Thanks for any
pointers.
-- 
 - - - - - - - - - - - - - - - - - - - - - - - -
Resistance is futile. Capacitance is efficacious.
Will Ware	email:    wware @ world.std.com



From boncelet at udel.edu  Thu Nov 25 08:02:27 1999
From: boncelet at udel.edu (Charles Boncelet)
Date: Thu, 25 Nov 1999 13:02:27 +0000
Subject: PythonMagick?
Message-ID: <383D3363.5312B2B7@udel.edu>

Does PythonMagick (a python interface to ImageMagick) exist anywhere?

I'd like to try it out, but the link to it is dead. (Start at
http://starship.skyport.net/crew/zack/, then click on PythonMagick, then
click again in the Installation area. But the file does not exist
there.)

Email to the author doesn't work either.

	Charlie Boncelet
-- 
Charles Boncelet <boncelet at udel.edu>
University of Delaware
Newark DE 19716 USA
http://www.eecis.udel.edu/~boncelet/



From fredrik at pythonware.com  Fri Nov 12 10:09:58 1999
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Fri, 12 Nov 1999 16:09:58 +0100
Subject: Tkinter and after
References: <382C2854.E9494EFE@lri.fr> <14380.11058.151680.518510@weyr.cnri.reston.va.us>
Message-ID: <00a001bf2d1f$fcd3ac70$f29b12c2@secret.pythonware.com>

Fred L. Drake, Jr. <fdrake at acm.org> wrote:
>  > It is said in Tkinter doc that after() is like a call to time.sleep()
>  > wich blocks Tkinter.
> 
>   I'm not sure what docs you're referring to (presumably the
> eff-bot's), but after() does not block Tk.

what version are you using?  if you
run this script, can you move the
scale widget after you've clicked
the button?

from Tkinter import *

root = Tk()

Scale(root).pack()

Button(root, text="click me!",
       command=lambda: root.after(100000)
       ).pack()

mainloop()

</F>




From mikl at linux-france.org  Mon Nov 29 14:12:37 1999
From: mikl at linux-france.org (Mickael Remond)
Date: Mon, 29 Nov 1999 19:12:37 +0000
Subject: Python meta object question
Message-ID: <3842D025.3283C871@linux-france.org>

Hello,

Is there a way to automatically update the class tree to show the change
in inherited class attributes after a class redefinition ?


Look at this example:

----------------------------------------
# Step one
# First description
class company1:
    company = "Company1"

class company2:
    company = "Company2"

class localcompany(company1):
    localcompany = "STE France"

class employee(localcompany):
    employee = "Me"

print
print "Step 1:"
print localcompany.company         # => Company1
print employee.company             # => Company1

# Step two
# localcompany is sold to company2
class localcompany(company2):
    localcompany="New STE France"

print
print "Step 2:"
print localcompany.company         # => Company2
print employee.company             # => Company1 (outdated!)

# Step three
# Force employee class update (same class definition)
class employee(localcompany):
    employee = "Me"

print
print "Step 3:"
print localcompany.company         # => Company2
print employee.company             # => Company2
-------------------------------------------------------------

This case is very simple but when the class tree becomes complicated it
is difficult to keep all class up to date.

Is there a way to redefined only one class, as in step 2 and have all
the other class automatically take this change into account ?

Thank you in advance for your help.

Mickael Remond



From T.Muddletyn at news.vex.net  Tue Nov 16 02:11:08 1999
From: T.Muddletyn at news.vex.net (T. Muddletyn)
Date: 16 Nov 1999 07:11:08 GMT
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <14384.16479.836502.161209@anthem.cnri.reston.va.us> <80pifh$12n1$1@hub.org> <8s_X3.676$pw2.12189912@news.randori.com>
Message-ID: <80r02c$1ho1$1@hub.org>

On Mon, 15 Nov 1999 12:55:58 -0800, Robin Dunn <robin at alldunn.com> wrote:
>Could it be moved to the Starship or perhaps to a machine at CNRI to be part
>of python.org?

Thanks to Gordon McMillan's initiative I have heard that Christian is willing
to host Parnassus somewhere on the Starship. This may be a big help. However,
if possible, i'd like to stay where I am, as I've have my little vex shell for
quite a while and am comfortable and familiar with it. However due to various
concerns about load/bandwidth raised here, I am discussing these things with
the vex.net admin to see what he thinks of all this. 

(ps. congrats on the new wxPython release today. wxPython is great... I'm
hoping to build myself some tools---love that listctrl <wink>---using it
eventually if i ever can manage to get around to unravelling its mysteries.)

...



From guido at cnri.reston.va.us  Tue Nov  2 08:47:17 1999
From: guido at cnri.reston.va.us (Guido van Rossum)
Date: 02 Nov 1999 08:47:17 -0500
Subject: __str__ vs. __repr__
References: <1270634443-26149069@hypernet.com> <000701bf24ef$e2d29b20$0f2d153f@tim> <19991102073422.B134170008@vislab.epa.gov>
Message-ID: <5lln8ht38a.fsf@eric.cnri.reston.va.us>

Randall Hopper <aa8vb at yahoo.com> writes:

> Tim Peters:
>  |The real illness is that lists (and dicts, and tuples) don't pass
>  |str-vs-repr'ness *down*.  That is, even though the example explicitly
>  |asks for str of a list, the list object asks for the repr of its
>  |elements.  That's convoluted: if (as happens to be true) str is meant to
>  |give a friendly string, why do the builtin container types ask their
>  |containees to produce unfriendly strings regardless? 
> 
> Thanks.  That's the source of my confusion.  I wanted to follow the
> language convention.  But Python's internals currently don't follow it, so
> the point is somewhat moot.

Actually, Python's internals *do* follow the convention.  Dicts and
lists don't define __str__(), so str() for them defaults to
__repr__(), and then of course repr() is used for the items.  It may
be confusing, it may not be what you want, but it *is* consistent ;-(

> For consistency, would it make sense to change this for Python 1.5.3 (that
> is, have sequence and dict types pass 'str-vs-repr'ness down)?

This has been asked a few times before (and usually it's been reported
as a bug, which it isn't -- see above).  I happened to see this post
and it made me delve deep into my intuition trying to figure out why I
don't like propagating str() down container items.

Here's what I think it is.  There's no reason why an object's str()
should be particularly suited to being included in list syntax.  For
example, I could have a list containing the following items:

	1		# an integer
	'1'		# a string (of an integer literal)
	'2, 3'		# a string containing a comma and a space
	'], ['		# a string containing list delimiters

Under the proposed rules, this list would print as:

	[1, 1, 2, 3], []

I would find this confusing and I worry that it could be used to fool
the user.

--Guido van Rossum (home page: http://www.python.org/~guido/)



From T.Moddletyn at news.vex.net  Mon Nov 15 12:42:56 1999
From: T.Moddletyn at news.vex.net (T. Moddletyn)
Date: 15 Nov 1999 17:42:56 GMT
Subject: Proposal: Python Info Collective
References: <dlBX3.2568$s87.54963@dfiatx1-snr1.gtei.net> <3d1z9sq2uc.fsf@amarok.cnri.reston.va.us> <JlFX3.3659$s87.67930@dfiatx1-snr1.gtei.net> <80oqn9$ocq$2@newshost.accu.uu.nl> <oLUX3.286$yA4.4821@dfiatx1-snr1.gtei.net>
Message-ID: <80pgn0$tnp$1@hub.org>

On Mon, 15 Nov 1999 14:27:32 GMT, Jules Allen wrote:
>The modest, retiring author has given me a little more insight in to the

This part is accurate. (-:

>architecture -- it's a low end machine using PyApache and he's built some

I don't know if i'd call it a low-end machine. But then again it's a long
time since i got to a computer store...

>kind of caching mechanism that's updated once a day. Mainly so the database

That's only the main page that is currently cached. I may generalise
the code to cache all the directories later though to improve 
performance if it seems necessary. It's not something that I really
want to do however. A "live" database is so much warmer and fuzzier
to cuddle up to. 

Thanks for the encouragement and help!




From gmcm at hypernet.com  Tue Nov  2 14:10:20 1999
From: gmcm at hypernet.com (Gordon McMillan)
Date: Tue, 2 Nov 1999 14:10:20 -0500
Subject: Comparison between Python and "Ruby"
In-Reply-To: <87zowwvl4t.fsf@ev.netlab.co.jp>
Message-ID: <1270552684-31022246@hypernet.com>

Yukihiro Matsumoto wrote:

> "Fredrik Lundh" <fredrik at pythonware.com> writes:
> 
> |> I guess he loves reducing variablity, if possible, which is
> totally |> against TMTOWTDI.  I don't judge whether it is good or
> bad. | |you're confusing syntax with semantics.
> 
> I don't think so.  I don't believe anyone can reduce semantic
> variability.  Thus TMTOWTDI applies to semantics AND syntax.

Did you mean to write what you wrote? From your premise, I 
would reach the opposite conclusion. Or are you saying that 
Guido also failed to reduce syntactic variability?

The problem with TMTOWTDI as used by Perlers is that they 
assume that because it is rare in Python for a given syntactic 
element to have an alternate form, that all successful attempts 
to solve a problem will have the same form.

A few years ago there were 19 different solutions posted to the 
problem of removing duplicates from a list, employing at least 
a half dozen distinctly different algorithms.
 
The same contest on c.l.pl.* might yield hundreds of 
solutions, but I doubt that the underlying algorithms would 
expand beyond that half dozen.

So I'd tend to agree with "I don't believe anyone can reduce 
semantic variability", but I would conclude that there's no 
connection between the syntactic and semantic levels.

> But it's OK.  I'm not Larry Wall nor Perl freak.
> 
> I just felt weird when Pythoneers say about TMTOWTDI or even
> TMTOWTHMTOWTDI. 
>       matz.
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list



- Gordon



From uhuenemoerder at hotmail.com  Mon Nov  8 16:41:59 1999
From: uhuenemoerder at hotmail.com (Ulrich Huenemoerder)
Date: Mon, 8 Nov 1999 22:41:59 +0100
Subject: PyRun_File / PyRun_String
References: <801vbn$2l64$1@bnews.gigabell.net>
Message-ID: <807dns$19hg$1@bnews.gigabell.net>

Hello again,


if somebody wants to know, this is how I solved it....

(I found out I have to use compiler option /MD using Visual C++ 6.0, too,
else the program crashed)

Ulrich.


#
# /my/script/path/myTest.py
#
def doPrint():
    print "called doPrint"

if __name__ == '__main__':
    print "this is __main__"

# end of myTest.py

//
// /somewhere/else/main.cpp
//
void main(int argc, char **argv) {

    /* variables */
    PyObject *pdict;

    Py_Initialize();

    pdict = PyDict_New();
    PyDict_SetItemString(pdict, "__builtins__", PyEval_GetBuiltins());

    // extend the searchpath
    PyRun_String("import sys", file_input, pdict, pdict);
    PyRun_String("sys.path.insert(0, '/my/script/path')", file_input, pdict,
pdict);

    // import the script -> no need for PyRun_File !!!
    PyRun_String("import myTest", file_input, pdict, pdict);

    // call the function defined in myTest.py
    PyRun_String("myTest.doPrint1()", file_input, pdict, pdict);
}

// end of main.cpp





From greg.ewing at compaq.com  Tue Nov 16 09:44:18 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Wed, 17 Nov 1999 03:44:18 +1300
Subject: Object Persistence and Pickle&zlib
References: <382BD762.9E2502AA@bibsyst.no>
Message-ID: <38316DC2.74E32B46@compaq.com>

thomas at bibsyst.no wrote:
> 
> class some_object:
>     title = ''
>     id = 0
>     items1 = {}
>     items2 = {}
>     items3 = {}
>     items4 = {}

That's going to give you four dictionaries stored as class attributes,
which are shared between all instances -- probably not what you want.
To give each instance its own set of dictionaries you'll have to
create them in the initialisation method:

   class some_object:
      title = ''
      id = 0

      def __init__(self):
         self.items1 = {}
         self.items2 = {}
         self.items3 = {}
         self.items4 = {}

That's also the reason your dictionaries weren't getting pickled.
Pickling an instance only saves instance attributes, not class
attributes.

> Some of the methods
> are available, but some are missing.

If you're reading the pickled data using a different program from
the one you wrote it with, check that you've defined all the
methods in the second program. Methods, being class attributes,
don't get pickled either -- it's up to the unpickling program
to supply them.

Hope that helps,
Greg



From aahz at netcom.com  Tue Nov  2 16:42:23 1999
From: aahz at netcom.com (Aahz Maruch)
Date: 2 Nov 1999 21:42:23 GMT
Subject: commas vs. spaces. vs. the legions of new lines
References: <XFMail.991102144322.mikael@isy.liu.se>
Message-ID: <7vnlrv$b98$1@nntp2.atl.mindspring.net>

In article <XFMail.991102144322.mikael at isy.liu.se>,
Mikael Olofsson  <mikael at isy.liu.se> wrote:
>
>standards-are-great--everyone-should-have-one-ly y'rs

"The wonderful thing about standards is that there are so many to choose
from."
--
                      --- Aahz (@netcom.com)

Androgynous poly kinky vanilla queer het    <*>      http://www.rahul.net/aahz/
Hugs and backrubs -- I break Rule 6  (if you want to know, do some research)



From MortenDaughtry at news.vex.net  Wed Nov 10 01:34:53 1999
From: MortenDaughtry at news.vex.net (Morten Daughtry)
Date: 10 Nov 1999 06:34:53 GMT
Subject: Writing dates to a ODBC-database
References: <38282605.3CD233E9@boxer.no>
Message-ID: <80b3md$2ia2$1@hub.org>

On Tue, 09 Nov 1999 14:47:49 +0100, Frank Biringer <fbi at boxer.no> wrote:
>distribution. I guess I have to create a dbi.dbiDate object. But since
>there's neither source or documentation for the dbi-module I don't know

python dbi documentation can be found here...
http://www.python.org/topics/database/





From ivanlan at callware.com  Mon Nov  8 12:44:39 1999
From: ivanlan at callware.com (Ivan Van Laningham)
Date: Mon, 08 Nov 1999 10:44:39 -0700
Subject: pythonistas or pythoneers?
References: <Pine.LNX.4.10.9911081026470.4450-100000@robert.polytopic.com> <slrn82e4mp.43f.thor@localhost.localdomain>
Message-ID: <38270C07.7BFAB861@callware.com>

Hi All--

Manuel Gutierrez Algaba wrote:
> 
> On Mon, 8 Nov 1999 10:29:01 -0600 (EST), Robert Meegan <rmeegan at murl.com> wrote:
> >On Sun, 7 Nov 1999, Juergen A. Erhard wrote:
> >
> >>
> >> I'm confused: are we pythoneers or pythonistas?
> 
> I'm quite amazed with these new word : "pythonistas"
> 
> Because it sounds too much Spanish. You see, here ( Spain),
> the football team supporters usually bear the suffix -istas.
> So we have Madridistas ( Real Madrid), Barcelonistas ( FC Barcelona)
> or Sevillistas ( FC Sevilla).
> 
> So pythonistas sounds extremely powerfull, because python can
> be spelled in Spanish as "piton", that means "horn" ( like the bull's
> horns). So pythonistas sounds like the "damned crazy followers
> of the horns" :DD

<g>
Are we not all "damned crazy followers of the horns" ??  Or at least
that crazy Dutchman?

<quod-erat-demonstrandum>-ly y'rs,
Ivan
----------------------------------------------
Ivan Van Laningham
Callware Technologies, Inc.
ivanlan at callware.com
ivanlan at home.com
http://www.pauahtun.org
See also: 
http://www.foretec.com/python/workshops/1998-11/proceedings.html
Army Signal Corps:  Cu Chi, Class of '70
Author:  Teach Yourself Python in 24 Hours
----------------------------------------------



From jsight at mindspring.com  Tue Nov 30 00:10:27 1999
From: jsight at mindspring.com (Jesse D. Sightler)
Date: Tue, 30 Nov 1999 00:10:27 -0500
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <81bmmi$279$1@nnrp1.deja.com> <3841EE77.6B5D6AC9@wjk.mv.com> <FLxz96.IEI@world.std.com>
Message-ID: <38435C43.B238EC15@mindspring.com>

Will Ware wrote:
> 
> William J. King (wjk at wjk.mv.com) wrote:
> : 4. conversion of integer to a string ( which may exist but havn't found
> : yet)
> 
> Try this:
>         str(123)
> or:
>         hex(123)
> 
> : 2. would like something like  'increment'/'decrement' operators
> :         either in form:
> :                     "x++"  or "incr x" vice "x = x + 1 " but its ok
> 
> I find myself grumbling about having to type "x = x + 1". The really
> clean thing to do, given that integers are objects, would be to define
> increment and decrement methods, so you'd type something like "i.incr()".
> Then you'd want to go around looking for other objects for which

Wouldn't this be impossible, since Integers are also immutable?  So, eg,
i.incr() could only return an incremented version of i, but not actually
increment i?

-- 
---------------
Jesse D. Sightler
http://www.biddin.com/



From phd at phd.russ.ru  Tue Nov 16 07:20:55 1999
From: phd at phd.russ.ru (Oleg Broytmann)
Date: Tue, 16 Nov 1999 12:20:55 +0000 (GMT)
Subject: Alright, I give up...chalk this one up to experience...
In-Reply-To: <383145c0.4445469@212.225.36.99>
Message-ID: <Pine.LNX.4.20.9911161220060.264-100000@fep132.fep.ru>

On Tue, 16 Nov 1999, Endgamer wrote:
> 
> olddate=os.popen("date",'r')
> olddate=readdate
> print olddate

   Why do you set olddate 2 times? What is "olddate=readdate"?

Oleg.
---- 
    Oleg Broytmann      Foundation for Effective Policies      phd at phd.russ.ru
           Programmers don't die, they just GOSUB without RETURN.




From htrd90 at zepler.org  Sun Nov 28 17:44:08 1999
From: htrd90 at zepler.org (Toby Dickenson)
Date: Sun, 28 Nov 1999 22:44:08 +0000
Subject: Python and COM: a bug?
References: <816010E2456BD111A48700805FBBE2EEFDF831@ex-quebec-u1.baan.com>
Message-ID: <Puo=OIr6Wof+GYnxsYrc7YwdPjqH@4ax.com>

Gaetan Corneau <Gaetan_Corneau at baan.com> wrote:

>fromwin32com.client.dynamic import Dispatch
>
>am = Dispatch("ITK.AdapterManager")
>ad = am.GetAdapter("OLEDB")
>Conn = ad.NewConnection()
>Conn.Open("/dsn:HPD380 /user:gc /pwd:qwerty")
>
>for Employee in Conn.Classes("Employee").GetObjects():
>    print Employee.Name
>
>
>
>What do we use all that for? In an integration server that manipulates
>abstract data, reading from a source, writing to a destination. The server
>lets you define data magppings graphically (for simple one to one field
>mapping). All that stuff has nothing to do with COM. The objects containing
>the data are C++ object. But when the mapping is complex, we create a COM
>wrapper over those C++ object (ITKDataObject) and send them to a script
>using ActiveX Scripting.
>
>So, the problem is this. When I access "Employee.Name", I get the ID for the
>Name property of Employee, but the Python COM caches it for ITKDataObject

Not quite. Clients are only allowed to cache the ID for use with that
one object. DISPIDs are allowed to be different between instances of
one class, just not during the lifetime of one object.

>From what you have described (if I understand correctly) then you
should be ok????????



>> ob = pythoncom.CoCreateInstance("Your.ProgID", None, pythoncom.CLSCTX_ALL,
>pythoncom.IID_IDispatch)
>> 
>> dispid = ob.GetIDsOfNames("foo")
>> print dispid
>> 
>> dispid2 = ob.GetIDsOfNames("foo")
>> print dispid2
>> 
>> if dispid != dispid2:
>>   print "Eeek"

ob2 = pythoncom.CoCreateInstance(...as above..)

dispid3 = ob2.GetIDsOfNames("foo")
print dispid3
if dispid != dispid3:
   print "Not so strange"



>> If it is the problem, you are clearly breaking the rules, and there is
>> a good chance that later versions of VBScript, or indeed any other COM
>> compatible language would also break.
>
>Now, this is a potentially severe bug. For the time being, the integration
>server is "locked" so all scripts must be written in VBScript, but it may
>not be so in future version. Another thing that could go wrong is that there
>is a new version of VBScript available. I haven't tried it yet, but I will:
>maybe this version uses caching.

iirc, the dispinterface proxy caches DISPIDs too. You might see the
problem even with VBScript if there is marshalling involved :-(



Toby Dickenson
htrd90 at zepler.org



From python-list at teleo.net  Wed Nov 10 01:10:50 1999
From: python-list at teleo.net (Patrick Phalen)
Date: Tue, 9 Nov 1999 22:10:50 -0800
Subject: Java vs. Python - Serialization Bakeoff
In-Reply-To: <mD6W3.25292$Ec1.280622@news1.frmt1.sfba.home.com>
References: <99110918085909.01567@quadra.teleo.net> <mD6W3.25292$Ec1.280622@news1.frmt1.sfba.home.com>
Message-ID: <9911092212380B.01567@quadra.teleo.net>

[claudius at catlover.com, on Tue, 09 Nov 1999]:
:: Patrick Phalen <python-list at teleo.net> says:
:: >http://starship.python.net/crew/gandalf/DNET/serial/jp-serial.html
:: >contains an report on tests comparing Java's and Python's data
:: >marshalling, using various criteria. Numerous interesting charts and
:: >graphs are presented.
:: >
:: >Since this was published in 1996 (Python 1.4), I wonder if anyone
:: >expert in this subject has any comment on whether the intervening years
:: >would have changed the conclusions substantively -- namely that Python
:: >is, on average, faster, simpler, and more flexible.
:: 
:: Hmm?  I get 'not found' (and I'm very interested in reading your findings.)

Ack! Apologies ... it's '.htm'




From coventry at room.mathdom.anu.edu.au  Wed Nov 24 23:40:58 1999
From: coventry at room.mathdom.anu.edu.au (Alex Coventry)
Date: 25 Nov 1999 15:40:58 +1100
Subject: automatic "lotjes trekken"
References: <81cqnv$2ip$1@vvs.superst.iae.nl> <bvwu2mddcnd.fsf@room.mathdom.anu.edu.au> <81hta3$1p7$1@vvs.superst.iae.nl>
Message-ID: <bvwr9hfchet.fsf@room.mathdom.anu.edu.au>

> I think this doesn't account for me not being allowed to draw myself.

Oh, yeah, perhaps Q should be chosen from R_{i-1}\P, and the
probabilities normalized accordingly.  Sorry.

Alex.



From mehta at mama.indstate.edu  Wed Nov  3 11:06:32 1999
From: mehta at mama.indstate.edu (mehta at mama.indstate.edu)
Date: Wed, 03 Nov 1999 16:06:32 GMT
Subject: assert module search
References: <000601bf25be$0cdeb880$2e2d153f@tim> <381FEF10.3B27C05B@webone.com.au>
Message-ID: <7vpmi4$1gp$1@nnrp1.deja.com>

which module does assert function belong to?

I tried to link to it through api docs but led me to exceptions module.


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



From mwh21 at cam.ac.uk  Thu Nov 25 03:28:46 1999
From: mwh21 at cam.ac.uk (Michael Hudson)
Date: 25 Nov 1999 08:28:46 +0000
Subject: Python complaints
References: <Bnx*fEKdo@news.chiark.greenend.org.uk> <867lj85g4c.fsf@g.local> <383CC835.B585EBDD@Lugoj.Com>
Message-ID: <m3ogcjrn41.fsf@atrus.jesus.cam.ac.uk>

James Logajan <JamesL at Lugoj.Com> writes:

> Sorry, but I think the price paid for "true" GC is too high for the
> small number of programs that benefit from this. Perhaps one should
> learn to clean up after oneself when cycles are expected.

While I'm inclined to agree that the loss of simplicity accompanied by
inclusion of a true GC is too high a price to pay, "Perhaps one should
learn to clean up after oneself when cycles are expected" is a
ridiculous statement.

Suppose I want to represent some chain like structure where I can
reach any element from any other. This is not an outrageous request,
to my mind. Clearly the easiest solution is to use some kind of doubly
linked list, but the na?ve implementation of this will lead to a
cyclic structure. So then either (a) you have to supply an explicit
.finalize() method or (b) you have to hack around mightily using
__getattr__ producing a solution that is slow and fragile. Neither of
these is at all satisfactory to my mind. Using DC's Aquisition classes
can help, but it's still an advanced technique that required to solve
a problem that looks incredibly simple.

Does anyone know if anything has come of that plan to have a rarely
running GC on dicts?

Regards,
Michael



From sblakey at u.washington.edu  Fri Nov 12 12:30:12 1999
From: sblakey at u.washington.edu (S. Blakey)
Date: Fri, 12 Nov 1999 09:30:12 -0800 (PST)
Subject: DBF interface
In-Reply-To: <382C38AF.69FFACF4@obop.com.pl>
Message-ID: <Pine.A41.4.10.9911120925480.47344-100000@dante03.u.washington.edu>

On Fri, 12 Nov 1999, [iso-8859-2] Przemys?aw G. [iso-8859-2] Gawro?ski wrote:

> Date: Fri, 12 Nov 1999 16:56:32 +0100
> From: "[iso-8859-2] Przemys?aw G. [iso-8859-2] Gawro?ski" <gawron at obop.com.pl>
> To: "python-list at python.org" <python-list at python.org>
> Subject: DBF interface
> 
> Is there a DBF file interface module ?
> 
> Thanks
> 
> Przemek
> 
> --
> Przemyslaw G. Gawronski UIN:8358522
> mailto:gawronskip at usa.net
> mailto:gawron at obop.com.pl
> 
> 
> 
> -- 
> http://www.python.org/mailman/listinfo/python-list
> 
Depends on what you mean by "DBF file"  Michal Spalinski wrote
 a module to read DB2 files (don't have the URL handy).  I have a slightly
modified version that can read MS Visual FoxPro Free tables - mail me if
you want it.  If you want to be able to write .dbf files - I don't know of
any existing module to do that, but it shouldn't be too hard.




From ivnowa at hvision.nl  Tue Nov 16 01:18:33 1999
From: ivnowa at hvision.nl (Hans Nowak)
Date: Tue, 16 Nov 1999 07:18:33 +0100
Subject: list.without()?
In-Reply-To: <y0jln7zi9pb.fsf@vier.idi.ntnu.no>
Message-ID: <199911160617.HAA07339@axil.hvision.nl>

On 16 Nov 99, Magnus L. Hetland wrote:

> 
> Hi!
> 
> Is there a function or method in the standard language that can
> non-destructively return a list without a specified element? Would it be
> possible to add it as a method to lists?
> 
> i.e:
> 
> # pseudocode
> 
> class list:
>     def without(self,element):
>         temp = self[:]
>         temp.remove(element)
>         return temp

Hmmz... Why a method?

def without(lst, elem):
    return filter(lambda x, elem=elem: x != elem, lst)


--Hans Nowak (zephyrfalcon at hvision.nl)
Homepage: http://fly.to/zephyrfalcon
You call me a masterless man. You are wrong. I am my own master.



From aa8vb at yahoo.com  Mon Nov 22 13:55:53 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Mon, 22 Nov 1999 13:55:53 -0500
Subject: Building wxPython [was Re: wxPython and DISLIN work nicely together]
In-Reply-To: <00e301bf32c6$e77aee90$0301a8c0@easystreet.com>
References: <80s0b4$g5n@mail.psy.uva.nl> <19991117090214.B807152@vislab.epa.gov> <017a01bf31f8$6bbea590$1a25d2d1@jenkondev.com> <19991119064944.A996886@vislab.epa.gov> <002301bf32b5$97a5d440$0301a8c0@easystreet.com> <19991119134422.A1004588@vislab.epa.gov> <00e301bf32c6$e77aee90$0301a8c0@easystreet.com>
Message-ID: <19991122135553.A1255406@vislab.epa.gov>

Robin Dunn:
 |This makes me wonder how you built wxWindows at all if these errors are in
 |the wxWindows headers...  What about the wxWindows sample apps, do they
 |compile?

Beats me but wxWindows worked out of the box.  configure, make, make
install.  I hadn't run any wxWindows test progs, but I just built/ran bombs
and it worked fine (minus a missing linker -rpath flag which some newer
toolkits don't get right for a while on Solaris, IRIX, FreeBSD, etc.).

BTW, I rebuilt and wxconfig --cflags looks as described now.

With the helpers.cpp NULL hack I just described, it got to the point of
linking wxcmodule.so, and failed trying to find wxPyHelpers:

    CC  -shared -all wx.o windows.o windows2.o windows3.o events.o misc.o misc2.o gdi.o mdi.o controls.o controls2.o cmndlgs.o stattool.o frames.o image.o printfw.o sizers.o clip_dnd.o -L/home/rhh/software/wxGTK-2.1.11/utils/wxPython/src `wx-config --libs`  -lwxPyHelpers   -o wxcmodule.so
    C++ prelinker: warning: could not locate library -lwxPyHelpers; assuming /usr/lib/libwxPyHelpers.a
    ld32: FATAL   9  : I/O error (-lwxPyHelpers): No such file or directory
    gmake: *** [wxcmodule.so] Error 2

build.py assumes that the utils/ directory is named wxPython.  I just left
it wxPython-2.1.11 since the README.txt file said "preferably...wxPython").

I'm also building under the wxGTK build tree's utils/ dir since I didn't
see whether wxPython expected to build under the wxGTK build or installed
trees, and under the build tree made more sense.  Judging from the default
search path of -L$(WXWIN)/utils/wxPython/src for files in the wxPython
build tree (WXWIN being the wxGTK installed tree), now I'm changing my
guess.  Which should it be?

Thanks,

Randall


-- 
Randall Hopper
aa8vb at yahoo.com



From darcy at vex.net  Mon Nov  8 06:17:22 1999
From: darcy at vex.net (D'Arcy J.M. Cain)
Date: 8 Nov 1999 11:17:22 GMT
Subject: Production Web Deployment?
References: <NBBBKMMANMGEEBHPJJECGEGIFJAA.dirkl@home.com>
Message-ID: <806bg2$2ob7$1@hub.org>

Dirk Leas <dirkl at home.com> wrote:
> I've read numerous threads regarding how to deploy Python on the web and
> there appears to be no real consensus on what actually performs/scales well
> (about 3M hits/month). Is anybody taking lots of hits with httpdapy,
> PyApache, ASP+Python, fastcgi, or Zope? I've got a customer who's
> development staff is struggling with the complexity of Java Servlets and
> they're already using a large 3rd party, vertical Python application (and
> several folks already use Python). Peliminary testing showed plain CGI was
> just too slow...

I use PyApache and I am pretty sure it won't scale well once I get
more hits.  I'm thinking about using httpdapy or perhaps even writing
my own Apache module with my own application built in.  The important
thing for me is database access.  Every page in my application opens
the database and that's quite wasteful.  I need a persistent database
connection.  What you need is hard to tell without knowing exactly
what you are trying to do.

I wouldn't just toss out Java altogether.  There are certain client
side things that Java (and JavaScript) are better suited for.  Use
the Python on the server side to generate Java and JavaScript pages
dynamically.  

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



From aa8vb at yahoo.com  Mon Nov 29 08:46:22 1999
From: aa8vb at yahoo.com (Randall Hopper)
Date: Mon, 29 Nov 1999 08:46:22 -0500
Subject: The First Annual Obfuscated Python Contest
In-Reply-To: <slrn83s2vv.1s0t.cjc26@tankgrrl.bridget.mindriot.net>
References: <mfbZ3.71$AZ6.1807@dfiatx1-snr1.gtei.net> <001d01bf328d$3ac77170$f29b12c2@secret.pythonware.com> <slrn83s2vv.1s0t.cjc26@tankgrrl.bridget.mindriot.net>
Message-ID: <19991129084622.A1839735@vislab.epa.gov>

Cliff Crawford:
 |Notice that all examples of obfuscated Python make liberal use of `lambda',
 |`map', and `reduce'.

Most use import too.  Hmmm...  ;-)

Randall



From dirk.engelmann at iwr.uni-heidelberg.de  Thu Nov  4 16:23:07 1999
From: dirk.engelmann at iwr.uni-heidelberg.de (Dirk Engelmann)
Date: Thu, 04 Nov 1999 16:23:07 -0500
Subject: debuging in shared library (using gdb or ddd)
Message-ID: <3821F93B.6BF2B62A@iwr.uni-heidelberg.de>

Hi!

How can I debug a shared library module imported into python

with gdb or ddd debugger ?

I would like to set breakpoints inside the (c++,c) shared library.

Thanks for help!

Dirk Engelmann

----------------------------------------------
Email: Dirk.Engelmann at iwr.uni-heidelberg.de






From ionel at psy.uva.nl  Fri Nov 12 13:07:49 1999
From: ionel at psy.uva.nl (Ionel Simionescu)
Date: Fri, 12 Nov 1999 19:07:49 +0100
Subject: docs
References: <80epm7$vbo$1@nnrp1.deja.com> <14378.61648.633649.431232@weyr.cnri.reston.va.us>
Message-ID: <80hkuv$oll@mail.psy.uva.nl>

Fred L. Drake, Jr. <fdrake at acm.org> wrote in message
news:14378.61648.633649.431232 at weyr.cnri.reston.va.us...
|
| mehta at mama.indstate.edu writes:
|  > I have noticed for the local docs of python in IE5 on winnt pages
|  > randomly donot get rendered but are left as raw html in browser as I go
|  > back and forth using browser buttons.  has any one experienced same
|
|   I have never seen this; if anyone can reproduce this and help
| isolate the cause I'd appreciate hearing about it.
|

Sorry, I've forgot to say that the bug does not bite the docs that come
packed for HTMLHelp.

Because they come from the same HTML sources, and because(*) internally both
HTMLHelp and IE use the same HTML parsing/rendering engine, I think that the
bug it's likely to be on the IE side.

ionel

---
(*) this is a guess, but I have little doubt about it.






From greg.ewing at compaq.com  Tue Nov  2 07:13:10 1999
From: greg.ewing at compaq.com (Greg Ewing)
Date: Wed, 03 Nov 1999 01:13:10 +1300
Subject: search (match) backward
References: <000301bf20f7$bd007d00$0a2d153f@tim> <3817CBE6.4A7BFD2B@mincom.com>
Message-ID: <381ED556.DF7A6E82@compaq.com>

John Farrell wrote:
> 
> However I believe for the subset of
> REs which do not use the constructs mentioned above, Tim's argument is
> correct.

Tim was talking about *real* REs. Perl's "regular" expressions are not
regular, and calling them such is a confusing abuse of terminology
(which we unfortunately seem to be stuck with).

Greg